Public flue exhaust control method and device and electronic equipment

Abstract

The invention discloses an exhaust control method and device for a public flue and electronic equipment. The method comprises the steps that the power measured value and the exhaust air rate measured value of a range hood of each floor in the public flue are obtained; establishing a causal effect model between the range hood power and the exhaust air rate according to the measured value of the range hood power and the measured value of the exhaust air rate; determining a to-be-optimized range hood associated with the target floor according to the causal effect model; and establishing an exhaust air rate optimization model based on the number of the to-be-optimized range hoods, and determining target power of the to-be-optimized range hoods according to the exhaust air rate optimization model. According to the method, the causal relationship between the range hood power of each floor and the exhaust air rate of the specific floor is analyzed through modeling, the range hood power is directionally controlled, the exhaust efficiency is optimized, energy waste is reduced, and the living comfort of the high-rise building is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of range hood control, and in particular to a common flue exhaust control method, a control device and an electronic device. Background Art

[0002] In modern high-rise buildings, the public flue system is an important facility that connects the kitchens of residents on each floor with the outdoor environment, and is responsible for discharging fumes and harmful gases generated during cooking. The effective operation of the public flue system is directly related to the quality of life of residents and the quality of indoor air.

[0003] In the existing public flue system, a fixed power control strategy, a timing control strategy or a pressure sensing control strategy is usually adopted to control the operation of the range hood. In the fixed power control strategy, the range hood operates according to the preset power, ignoring the usage differences between floors and the dynamic changes of actual exhaust demand, which not only leads to low exhaust efficiency, but also causes serious energy waste; in the timing control strategy, the range hood is turned on or off according to a fixed schedule, and the exhaust volume cannot be dynamically adjusted according to the actual exhaust demand, resulting in low exhaust efficiency; in the pressure sensing control strategy, the range hood power is adjusted according to the pressure change in the flue, but this method usually responds slowly and it is difficult to accurately control the exhaust volume of the range hood, affecting the exhaust efficiency and exhaust effect. Summary of the invention

[0004] The present invention provides a common flue exhaust control method, a control device and an electronic device to solve the problem that the existing common flue range hood control strategy cannot dynamically adjust the exhaust volume according to the usage differences between floors and the actual exhaust demand, resulting in low exhaust efficiency, and can optimize the exhaust efficiency.

[0005] According to one aspect of the present invention, a common flue exhaust control method is provided, comprising: obtaining the actual measured values ​​of range hood power and exhaust volume on each floor in the common flue; establishing a causal effect model between range hood power and exhaust volume according to the actual measured values ​​of range hood power and exhaust volume; determining the range hood to be optimized associated with the target floor according to the causal effect model; establishing an exhaust volume optimization model based on the number of range hoods to be optimized, and determining the target power of the range hood to be optimized according to the exhaust volume optimization model.

[0006] Optionally, a causal effect model between range hood power and exhaust volume is established based on the actual measured values ​​of range hood power and exhaust volume, including: taking the actual measured values ​​of range hood power of each floor except the characteristic floor as input parameters, and taking the predicted values ​​of range hood power of the characteristic floor as output parameters, to establish a processing variable prediction model; taking the actual measured values ​​of range hood power of all floors as input parameters, and taking the predicted values ​​of exhaust volume of the target floor as output parameters, to establish a result variable prediction model; and establishing a causal effect model based on the processing variable prediction model and the result variable prediction model.

[0007] Optionally, a causal effect model is established based on the processing variable prediction model and the result variable prediction model, including: calculating the processing variable residual based on the actual measured value of the range hood power on the characteristic floor and the predicted value of the processing variable prediction model; calculating the result variable residual based on the actual measured value of the exhaust volume on the target floor and the predicted value of the result variable prediction model; and establishing the causal effect model based on the processing variable residual and the result variable residual.

[0008] Optionally, a causal effect model between range hood power and exhaust volume is established based on the actual measured values ​​of range hood power and exhaust volume, including: establishing a training set and a test set based on the actual measured values ​​of range hood power and exhaust volume; cross-validating the causal effect model based on the training set and the test set until the causal effect model meets the predictive performance conditions.

[0009] Optionally, an exhaust volume optimization model is established based on the number of range hoods to be optimized, and the target power of the range hoods to be optimized is determined according to the exhaust volume optimization model, including: taking the range hood power of the range hood to be optimized as an optimization parameter, and taking the exhaust volume prediction value of the target floor as the optimization result, to establish the exhaust volume optimization model; and / or, taking the measured value of the range hood power of each floor except the range hood to be optimized as a fixed parameter, taking the range hood power of the range hood to be optimized as an optimization parameter, and taking the exhaust volume prediction value of the target floor as the optimization result, to establish the exhaust volume optimization model.

[0010] Optionally, obtaining the actual measured values ​​of range hood power and exhaust volume on each floor in the public flue includes: executing at least one of the following data processing strategies on the actual measured values ​​of range hood power and exhaust volume: data cleaning, data denoising and data standardization.

[0011] Optionally, the public flue exhaust control method further includes: acquiring environmental parameters of each floor in the public flue; and optimizing the causal effect model according to the environmental parameters.

[0012] Optionally, the causal effect model is optimized according to the environmental parameters, including: taking the measured values ​​of the range hood power and the environmental parameters of each floor except the characteristic floor as input parameters, and taking the predicted values ​​of the range hood power of the characteristic floor as output parameters, to establish an optimized processing variable prediction model; taking the measured values ​​of the range hood power and the environmental parameters of all floors as input parameters, and taking the predicted values ​​of the exhaust volume of the target floor as output parameters, to establish an optimized result variable prediction model; and establishing a causal effect model based on the optimized processing variable prediction model and the result variable prediction model.

[0013] According to another aspect of the present invention, a common flue exhaust control device is provided, comprising: a data acquisition module for acquiring the actual measured values ​​of range hood power and exhaust volume on each floor in the common flue; a cause-effect analysis module for establishing a cause-effect effect model between range hood power and exhaust volume according to the actual measured values ​​of range hood power and exhaust volume; a calculation module for determining the range hood to be optimized associated with the target floor according to the cause-effect effect model; and an optimization module for establishing an exhaust volume optimization model based on the number of range hoods to be optimized, and determining the target power of the range hood to be optimized according to the exhaust volume optimization model.

[0014] According to another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned common flue exhaust control method.

[0015] The technical solution of the embodiment of the present invention establishes a causal effect model between range hood power and exhaust volume by acquiring the actual measured values ​​of range hood power and exhaust volume on each floor in a public flue, and determines the range hood to be optimized associated with the target floor according to the causal effect model. After determining the range hood to be optimized, an exhaust volume optimization model for the target floor is established based on the number of range hoods to be optimized, and the target power of the range hood to be optimized is determined according to the exhaust volume optimization model. This solves the problem of low exhaust efficiency caused by the existing control strategy of range hoods in public flues, which cannot dynamically adjust the exhaust volume according to the usage differences between floors and the actual exhaust demand. It can respond to the actual exhaust demand of each floor in a timely manner, optimize the exhaust efficiency of the public flue, and reduce energy waste of range hoods.

[0016] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flow chart of a common flue exhaust control method provided by an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of the structure of a public flue provided in an embodiment of the present invention;

[0020] Figure 3 A flow chart of a modeling method of a causal effect model provided by an embodiment of the present invention;

[0021] Figure 4 A flow chart of another modeling method of a causal effect model provided by an embodiment of the present invention;

[0022] Figure 5 A flow chart of another common flue exhaust control method provided by an embodiment of the present invention;

[0023] Figure 6 A flow chart of another common flue exhaust control method provided by an embodiment of the present invention;

[0024] Figure 7 A schematic structural diagram of a common flue exhaust control device provided by an embodiment of the present invention;

[0025] Figure 8 It is a structural schematic diagram of an electronic device for implementing the common flue exhaust control method of an embodiment of the present invention.

[0026] Reference numerals:

[0027] 101. Data acquisition module; 102. Cause-and-effect analysis module; 103. Calculation module; 104. Optimization module; 10. Electronic device; 11. Processor; 12. ROM (Read Only Memory); 13. RAM (Random Access Memory); 14. Bus; 15. I / O interface; 16. Input unit; 17. Output unit; 18. Storage unit; 19. Communication unit. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "target floor", "feature floor", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] Figure 1 A flow chart of a public flue exhaust control method provided in an embodiment of the present invention. This embodiment is suitable for application scenarios of power regulation of range hoods in a public flue system in a high-rise building.

[0031] Figure 2 A structural schematic diagram of a public flue provided in an embodiment of the present invention takes an L-story building (such as the 1st floor, ..., the i-th floor, the i+1th floor, ..., the L-th floor from bottom to top) as an example, and exemplarily shows the structure of a public flue in a high-rise building.

[0032] See also Figure 1 and Figure 2 As shown, the public flue exhaust control method of the present application specifically includes:

[0033] S1: Obtain the actual measured values ​​of range hood power and exhaust volume on each floor in the public flue.

[0034] The range hood power refers to the output power of the motor when the range hood performs the range hood extraction operation. In this embodiment, the actual measured value of the range hood power on each floor can be obtained through real-time communication with the range hood controller. Figure 2 As shown in the figure, the measured power value of the range hood on the first floor is recorded as P1, the measured power value of the range hood on the second floor is recorded as P2, ..., the measured power value of the range hood on the i-th floor is recorded as P i , ..., the measured power value of the range hood on the Lth floor is recorded as P L.

[0035] The exhaust volume represents the exhaust volume of each floor in the public flue. In this embodiment, a detection element (such as a flow sensor or a wind speed sensor) can be set at the connection position between the public flue and each floor to collect the actual measured value of the exhaust volume corresponding to each floor. Figure 2 As shown, the measured value of the exhaust volume of the first floor is recorded as Q1, the measured value of the exhaust volume of the second floor is recorded as Q2, ..., the measured value of the exhaust volume of the i-th floor is recorded as Q i , ..., the measured value of the exhaust volume of the Lth layer is recorded as Q L .

[0036] Optionally, obtaining the measured power value and exhaust volume value of the range hoods on each floor in the public flue includes: executing at least one of the following data processing strategies on the measured power value and the measured exhaust volume value: data cleaning, data denoising and data standardization. Among them, data cleaning refers to processing and screening the measured values ​​to ensure the accuracy, completeness and consistency of the data. Typically, data cleaning methods include but are not limited to: processing missing values, outliers, deleting duplicate values ​​and other steps. Data denoising refers to removing interfering data from the measured values, which may be caused by measurement errors, equipment failures or other external factors. Typically, data denoising methods include but are not limited to: mean filtering, median filtering, wavelet transform, connectivity graph method, support vector machine, etc. Data standardization refers to converting the measured values ​​from the original range to a specific range. Typically, data standardization includes but is not limited to: normalization.

[0037] Specifically, after the exhaust operation is started in the public flue, the power data of all range hoods in the high-rise building are collected, and the power data of each floor are respectively processed by data cleaning, data denoising or data standardization, and the finally obtained power data are recorded as the actual measured value of the range hood power of the corresponding floor; at the same time, the exhaust volume data of all floors are collected, and the exhaust volume data of each floor are respectively processed by data cleaning, data denoising or data standardization, and the finally obtained exhaust volume data are recorded as the actual measured value of the exhaust volume of the corresponding floor.

[0038] S2: Establish a causal effect model between range hood power and exhaust volume based on the actual measured values ​​of range hood power and exhaust volume.

[0039] The causal effect model is a model established based on the causal relationship between the range hood power of each floor and the exhaust volume of the target floor. In this embodiment, the causal effect model can be established based on double machine learning (DML) technology. The double machine learning technology is an algorithm that combines a machine learning model to perform causal relationship inference.

[0040] Specifically, in the causal effect model of the present application, the range hood power of each floor is used as a processing variable (i.e., the independent variable of the model), and the exhaust volume of the target floor is used as the result variable (i.e., the dependent variable of the model). A regression prediction model is established based on the residual between the predicted value and the measured value of the processing variable, and the residual between the predicted value and the measured value of the result variable. The model is optimized through machine learning to calculate the causal relationship between the range hood power of each floor and the exhaust volume of the target floor. The target floor is a floor that needs to be configured based on actual range hood control. In this embodiment, the target floor can be based on a specific control sequence or a user-defined configuration.

[0041] S3: Determine the range hood to be optimized associated with the target floor according to the causal effect model.

[0042] The range hood to be optimized refers to the range hood that has a relatively large impact on the exhaust volume of the target floor. Correspondingly, the range hood that has a relatively small impact on the exhaust volume of the target floor is defined as the range hood not to be optimized. In this embodiment, only the range hood power of the range hood to be optimized needs to be adjusted, while the range hood power of the range hood not to be optimized continues to remain unchanged.

[0043] Specifically, the measured value of the exhaust volume of the target floor can be substituted into the causal effect model, and the range hoods whose treatment variables have a greater impact on the degree of change of the outcome variable can be selected. For example, the range hood power of any floor can be substituted into the causal effect model trained in the above steps. If the model calculation result is greater than or equal to the impact threshold, the range hood on this floor is determined as the range hood to be optimized; if the model calculation result is less than the impact threshold, the range hood on this floor is determined as the range hood not to be optimized. In this embodiment, the change rate threshold can be established based on the calibration of the degree of change of the exhaust volume of the target floor caused by an increase of one unit in the range hood power, and its value is not limited.

[0044] For example, if a building where a public flue is located has L floors in total, and n range hoods to be optimized are calculated through the causal effect model, then the number of range hoods not to be optimized is Ln.

[0045] S4: establishing an exhaust volume optimization model based on the number of range hoods to be optimized, and determining the target power of the range hoods to be optimized according to the exhaust volume optimization model.

[0046] Among them, the exhaust volume optimization model is a regression prediction model established by taking the range hood power of the range hood to be optimized as the optimization parameter and the exhaust volume of the target floor as the optimization target.

[0047] In some embodiments, an exhaust volume optimization model is established based on the number of range hoods to be optimized, and the target power of the range hoods to be optimized is determined according to the exhaust volume optimization model, including: using the range hood power of the range hood to be optimized as an optimization parameter, using the exhaust volume prediction value of the target floor as the optimization result, and establishing the exhaust volume optimization model. Specifically, if a building where a public flue is defined has L floors, and n range hoods to be optimized are identified through the causal effect model, the power of the n range hoods to be optimized is used as the optimization parameter, recorded as p1', p2', ..., p n ', the predicted value of the exhaust volume of the target floor is taken as the optimization result, recorded as Q j ', then the exhaust volume optimization model is established as shown in Formula 1:

[0048]

[0049] Among them, y j (p'1,p'2,...,p' n ) represents the regression prediction function of the j-th layer exhaust volume based on n oil fume probabilities to be optimized; Q min Indicates the lower limit threshold of the range hood exhaust volume; P max Indicates the upper threshold of the range hood power.

[0050] In other embodiments, an exhaust volume optimization model is established based on the number of range hoods to be optimized, and the target power of the range hoods to be optimized is determined according to the exhaust volume optimization model, including: using the measured power values ​​of the range hoods on each floor except the range hoods to be optimized as fixed parameters, using the range hood power of the range hoods to be optimized as optimization parameters, and using the predicted exhaust volume value of the target floor as the optimization result to establish the exhaust volume optimization model. Specifically, if a building where a public flue is defined has L floors in total, and n range hoods to be optimized are identified through the causal effect model, the power of the n range hoods to be optimized is used as the optimization parameter, which is recorded as p1', p2', ..., p n ', the measured values ​​of the range hood power of m (L = m + n) range hoods to be optimized are taken as fixed parameters, denoted as g1, g2, ..., g m , the predicted value of the exhaust volume of the target floor is taken as the optimization result, recorded as Q j ', then the exhaust volume optimization model of the target floor is established as shown in Formula 2:

[0051]

[0052] Among them, y j (g1,g1,…,g m ,p'1,p'2,…,p' n ) represents the regression prediction function of the exhaust volume of the jth floor based on the power of the range hoods on all floors; Q minIndicates the lower limit threshold of the range hood exhaust volume; P max Indicates the upper limit of the range hood power

[0053] Specifically, after the exhaust operation is started in the public flue, the power data of all range hoods in the high-rise building and the exhaust volume data of all floors are collected, and the actual measured values ​​of the range hood power and the exhaust volume are recorded. A causal effect model is established based on the causal relationship between the range hood power of each floor and the exhaust volume of the target floor, and the aforementioned measured values ​​of the range hood power and the exhaust volume are substituted into the causal effect model for model training and testing. After the causal effect model training is completed, the measured value of the exhaust volume of the target floor is substituted into the causal effect model, and n range hoods to be optimized whose treatment variables have a greater impact on the degree of change of the outcome variable are selected. Referring to the above formula 1 or formula 2, an exhaust volume optimization model for the target floor is established, and combined with the genetic algorithm optimization method, the target power of the n range hoods to be optimized is found, and then the range hood power of the range hood to be optimized is controlled to reach the corresponding target power.

[0054] Therefore, through modeling and analysis of the causal relationship between the range hood power of each floor and the exhaust volume of a specific floor, the range hood power that has a greater impact on the exhaust volume of the target floor is controlled in a targeted manner, which solves the problem of low exhaust efficiency caused by the existing public flue range hood control strategy, which is unable to dynamically adjust the exhaust volume according to the usage differences between floors and the actual exhaust demand. This can respond to the actual exhaust demand of each floor in a timely manner, optimize the exhaust efficiency of the public flue, reduce energy waste, and improve the living comfort of high-rise buildings.

[0055] Figure 3 A flowchart of a modeling method of a causal effect model provided by an embodiment of the present invention, in Figure 1 Based on the embodiment shown, a specific implementation method of establishing a causal effect model is exemplarily shown. Figure 3 As shown, in the above step S2, a causal effect model between the range hood power and the exhaust volume is established according to the actual measured value of the range hood power and the actual measured value of the exhaust volume, which specifically includes the following steps:

[0056] S201: Taking the measured values ​​of the range hood power of each floor except the characteristic floor as input parameters, and taking the predicted values ​​of the range hood power of the characteristic floor as output parameters, a processing variable prediction model is established.

[0057] The characteristic floor is any floor in the building where the public flue is located. Figure 2 As shown in FIG. 1 , if the building where the public flue is located has a total of L floors, and the i-th floor is defined as the characteristic floor, the processing variable prediction model is based on the measured values ​​of the range hood power of each floor except the characteristic floor (for example, the i-th floor) (for example, P1, P2, ..., P i-1 , P i+1, ..., P L ) Calculate the predicted value P of the range hood power of the characteristic floor (for example, the i-th floor) i '. For example, taking the L-story building as an example, a processing variable prediction model as shown in Formula 3 can be established:

[0058] P i '=z i (W i )=z i (P1,P2,……,P i-1 ,P i+1 ,……,P L ),1≤i≤L (Formula 3)

[0059] Among them, z i (W i ) indicates that based on W i The range hood power regression prediction model for the characteristic floor (for example, the i-th floor) is established; W i represents the measured power values ​​of the range hoods on each floor except the characteristic floor (e.g., the i-th floor) (P1, P2, ..., P i-1 , P i+1 , ..., P L );P i ' represents the predicted value of the range hood power of the characteristic floor (for example, the i-th floor).

[0060] S202: The measured values ​​of the range hood power on all floors are used as input parameters, and the predicted value of the exhaust volume on the target floor is used as an output parameter to establish a result variable prediction model.

[0061] The target floor is a floor based on a specific control sequence or user-defined configuration. Figure 2 As shown in the figure, if the building where the public flue is located has L floors in total and the jth floor is defined as the target floor, the result variable prediction model is based on the measured values ​​of the range hood power on all floors (e.g. P1, P2, ..., P j-1 , P j , P j+1 , ..., P L ) Calculate the exhaust volume prediction value Q of the target floor (for example, the jth floor) j '. For example, taking the L-story building as an example, a result variable prediction model as shown in Formula 4 can be established:

[0062] Q' j =y j (W) = y j (P1,P2,……,P L ) (Formula 4)

[0063] Among them, y j(W) represents the range hood power regression prediction model of the target floor (for example, the jth floor) established based on W; W represents the measured values ​​of the range hood power of all floors (P1, P2, ..., P L );Q' j Indicates the predicted value of the exhaust volume of the target floor (for example, the jth floor).

[0064] S203: Establish a causal effect model based on the treatment variable prediction model and the outcome variable prediction model.

[0065] In this embodiment, the processing variable prediction model and the result variable prediction model are self-learning models established based on dual machine learning technology.

[0066] In some embodiments, a causal effect model is established based on a processing variable prediction model and a result variable prediction model, including: calculating the processing variable residual based on the actual measured value of the range hood power of the feature floor and the predicted value of the processing variable prediction model; calculating the result variable residual based on the actual measured value of the exhaust volume of the target floor and the predicted value of the result variable prediction model; and establishing a causal effect model based on the processing variable residual and the result variable residual.

[0067] Specifically, combining the above formulas 3 and 4, the residual of the treatment variable is equal to Zz i (W i ), the residual of the resulting variable is equal to Yy j (W), and solve the causal effect model θ(W) by optimizing the following formula 5 i ):

[0068] Yy j (W) = θ(W i )*[Zz i (W i )]+∈ (Formula 5)

[0069] Among them, ∈ represents random error; θ(W i ) represents the value based on the variable W i The causal effect function between the treatment variable and the structural variable is established, and its physical meaning is i Under the environment, the degree of change in the predicted value of the outcome variable Y (the exhaust volume of the j-th target floor) caused by an increase of one unit in the treatment variable Z (the measured value of the range hood power on the i-th characteristic floor).

[0070] Therefore, the technical solution of the present invention constructs a processing variable prediction model and a result variable prediction model based on a machine learning algorithm, and establishes a causal effect model between the range hood power of each floor and the exhaust volume of a specific floor based on the residual between the model prediction value and the measured value. It can accurately identify the floors that have a greater impact on the target floors, realize directional control of the range hoods, and avoid power regulation on all floors resulting in low power regulation accuracy.

[0071] Figure 4 A flowchart of another causal effect modeling method provided by an embodiment of the present invention, in Figure 1 Based on the embodiment shown, a specific implementation method of causal effect model training is exemplarily shown. Figure 4 As shown, in the above step S2, a causal effect model between the range hood power and the exhaust volume is established according to the measured value of the range hood power and the measured value of the exhaust volume, and further includes:

[0072] S204: Establish a training set and a test set based on the actual measured values ​​of the range hood power and the exhaust volume.

[0073] The data in the training set is used to train the causal effect model; the data in the test set is used to verify the causal effect model. In this embodiment, the measured values ​​of the range hood power and exhaust volume at the same sampling time on all floors form an associative array, and each training set at least includes: multiple associative arrays obtained by continuous sampling within a specific time period.

[0074] Specifically, based on any machine learning algorithm, data processing can be performed on the input features (such as the measured value of the range hood power) and the output features (such as the predicted value of the exhaust volume or the predicted value of the range hood power), such as using any standardization algorithm to perform data normalization processing on the input features (such as the measured value of the range hood power) and the output features (such as the predicted value of the exhaust volume or the predicted value of the range hood power) to make them suitable for the input / output requirements of the neural network model, and using clustering or averaging to divide the processed data into k sub-data sets. After determining the input features and output features, a regression model with one or more LSTM layers followed by fully connected layers (Dense layers) is constructed as a causal effect model. The input layer data dimension of the regression model is equal to the number of input features (such as the measured values ​​of the range hood power of all floors or the measured values ​​of the range hood power of all floors except the feature floors), and the output layer data dimension of the regression model is equal to the number of output features (such as the predicted value of the exhaust volume or the predicted value of the range hood power). Preferably, one or two LSTM layers are set in the causal effect model. Preferably, the number of neurons in any LSTM layer in the causal effect model is greater than or equal to 10 and less than or equal to 1024.

[0075] S205: Cross-validate the causal effect model based on the training set and the test set until the causal effect model meets the prediction performance condition.

[0076] The prediction performance condition is a judgment condition established based on the stability and accuracy of the prediction model. For example, when the output parameters of the prediction model converge and the residual between the model prediction value and the actual value is less than a preset residual threshold, the causal effect model is judged to meet the prediction performance condition.

[0077] Specifically, the processing variable prediction model and the outcome variable prediction model are cross-validated. In the cross-validation process, some arrays in the k sub-datasets are used as test sets and the remaining arrays are used as training sets. i (W i ), the input features are the measured values ​​of the range hood power of each floor except the feature floor (for example, the i-th floor) (for example, P1, P2, ..., P i-1 , P i+1 , ..., P L ), the output feature is the predicted value of the range hood power of the feature floor (for example, the i-th floor). i (W i ) After training is completed, evaluate the model's predictive performance, such as stability and accuracy. If the processing variable prediction model z i (W i ) is unstable or inaccurate, adjust the parameters of the machine learning algorithm and retrain until the processing variable prediction model z i (W i ) reaches the prediction performance condition. The model that finally reaches the prediction performance condition is used as the optimal processing variable prediction model to calculate the predicted value of the processing variable.

[0078] Accordingly, in the outcome variable prediction model y j In the training and test sets of (W), the input features are the measured values ​​of the range hood power on all floors (e.g., P1, P2, ..., P j-1 , P j , P j+1 , ..., P L ), the output feature is the predicted value of the exhaust volume of the target floor (for example, the jth floor). In the result variable prediction model y j (W) After training, evaluate the model’s predictive performance, such as stability and accuracy. If the outcome variable predicts the model y j If the prediction result of (W) is unstable or not accurate enough, adjust the parameters of the machine learning algorithm and retrain until the result variable prediction model y j(W) reaching the prediction performance condition, and taking the model that finally reaches the prediction performance condition as the best outcome variable prediction model, which is used to calculate the predicted value of the outcome variable.

[0079] Therefore, the technical solution of the present invention, by constructing a test set and a training set, cross-validates the causal effect model, which is beneficial to improving the stability and accuracy of the model and improving the control accuracy of exhaust regulation based on the causal effect model.

[0080] Figure 5 A flowchart of another common flue exhaust control method provided by an embodiment of the present invention. Figure 5 As shown, in combination with the above embodiments, the public flue exhaust control method of the present application specifically includes the following steps:

[0081] S501: Start.

[0082] S502: Collect exhaust volume and range hood power data on each floor.

[0083] S503: Perform data preprocessing.

[0084] In this embodiment, data preprocessing includes but is not limited to: data cleaning, data denoising and data standardization.

[0085] S504: Divide the preprocessed data into k sub-data sets.

[0086] S505: Initialize the regression prediction model.

[0087] In this embodiment, a regression model is established based on a machine learning model.

[0088] S506: Iterate cyclically based on k sub-data sets.

[0089] S507: Use part of the array as a test set and the rest of the array as a training set.

[0090] S508: Regress the prediction model on the training set, and evaluate the regression prediction model on the test set.

[0091] S509: Record the model evaluation results.

[0092] S510: Determine whether k iterations are completed.

[0093] If k iterations are completed, step S511 is executed; if k iterations are not completed, step S506 is executed.

[0094] S511: Determine whether the model is stable and accurate.

[0095] If the model is stable and accurate, step S513 is executed; if the model is unstable or inaccurate, step S512 is executed.

[0096] S512: Adjust the regression prediction model parameters.

[0097] After the model parameters are adjusted, the process returns to step S506.

[0098] S513: Determine the best regression prediction model.

[0099] S514: Retrain the regression prediction model.

[0100] S515: Obtain a regression prediction model of the exhaust volume of the target floor and the range hood power of each floor, and a regression prediction model of the range hood power of each floor and the range hood power of other floors.

[0101] S516: Calculate the residual between the predicted value and the measured value of the regression prediction model, and derive a causal effect model between the exhaust volume of the target floor and the power of the range hoods on each floor based on the residual.

[0102] S517: Determine the range hood to be optimized that is associated with the target floor.

[0103] S518: Determine optimization parameters and target parameters.

[0104] In this embodiment, the optimization parameter may be the range hood power of the range hood to be optimized, and the target parameter may be the exhaust volume of the target floor.

[0105] S519: Establish an exhaust volume optimization model.

[0106] S520: Execute an optimization algorithm.

[0107] S521: Obtain the target power of the range hood to be optimized.

[0108] S522: End.

[0109] Figure 6 A flowchart of another public flue exhaust control method provided in an embodiment of the present invention is provided. Figure 1 Based on the embodiment shown, the modeling accuracy of the causal effect model is optimized by introducing environmental parameters. Figure 6 As shown, the public flue exhaust control method of the present application includes the following steps:

[0110] S601: Obtain the actual measured values ​​of the range hood power and exhaust volume on each floor in the public flue.

[0111] S602: Establish a causal effect model between the range hood power and the exhaust volume according to the actual measured value of the range hood power and the actual measured value of the exhaust volume.

[0112] S603: Obtain environmental parameters of each floor in the public flue.

[0113] In this embodiment, the environmental parameters refer to parameters that affect the smoke exhaust airflow and smoke exhaust resistance. Typically, the environmental parameters include but are not limited to at least one of the following: atmospheric pressure, temperature and humidity.

[0114] S604: Optimize the causal effect model according to environmental parameters.

[0115] In this embodiment, the causal effect model is optimized according to the environmental parameters, including: using the environmental parameters as input parameters to train at least one prediction model in the causal effect model, such as a treatment variable prediction model or an outcome variable prediction model.

[0116] Optionally, the causal effect model is optimized according to the environmental parameters, including: taking the measured values ​​of the range hood power and the environmental parameters of each floor except the characteristic floor as input parameters, and taking the predicted values ​​of the range hood power of the characteristic floor as output parameters, to establish an optimized processing variable prediction model; taking the measured values ​​of the range hood power and the environmental parameters of all floors as input parameters, and taking the predicted values ​​of the exhaust volume of the target floor as output parameters, to establish an optimized result variable prediction model; and establishing a causal effect model based on the optimized processing variable prediction model and the result variable prediction model.

[0117] Specifically, see Figure 2 As shown in FIG. 1 , if the building where the public flue is located has L floors, the i-th floor is defined as the characteristic floor, and the environmental parameter is defined as η, then the processing variable prediction model after the environmental parameter optimization is as shown in Formula 6:

[0118] P i '=z i (W i )

[0119] =z i (P1,P2,……,P i-1 ,P i+1 ,……,P L ,η1,η2,……,η i-1 ,η i+1 ,……,η L ),1≤i≤L

[0120] (Formula 6)

[0121] Among them, z i (W i ) indicates that based on W i The range hood power regression prediction model for the characteristic floor (for example, the i-th floor) is established; W i represents the measured power values ​​of the range hoods on each floor except the characteristic floor (e.g., the i-th floor) (P1, P2, ..., P i-1 , Pi+1 , ..., P L ) and environmental parameters (η1,η2,……,η i-1 ,η i+1 ,……,η L );P i ' represents the predicted value of the range hood power of the characteristic floor (for example, the i-th floor).

[0122] The result variable prediction model based on environmental parameter optimization is shown in Formula 7:

[0123] Q' j =y j (W) = y j (P1,P2,……,P L ,η1,η2,……,η L ) (Formula 7)

[0124] Among them, y j (W) represents the range hood power regression prediction model of the target floor (for example, the jth floor) established based on W; W represents the measured values ​​of the range hood power of all floors (P1, P2, ..., P L ) and environmental parameters (η1,η2,……,η L );Q' j Indicates the predicted value of the exhaust volume of the target floor (for example, the jth floor).

[0125] Therefore, the technical solution of the present invention, by introducing environmental parameters, expands the dimensions of the model input parameters, reduces the influence of environmental parameter changes on the causal effect between the range hood power of each floor and the exhaust volume of a specific floor, optimizes the modeling accuracy of the causal effect model, and improves the control accuracy of exhaust regulation based on the causal effect model.

[0126] Based on the above-mentioned inventive concept, an embodiment of the present invention further provides a common flue exhaust control device, which can execute the common flue exhaust control method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0127] Figure 7 The following is a schematic diagram of the structure of a common flue exhaust control device provided by an embodiment of the present invention. Figure 7 As shown, the public flue exhaust control device of the present application includes: a data acquisition module 101, a cause-effect analysis module 102, a calculation module 103 and an optimization module 104.

[0128] Among them, the data acquisition module 101 is used to obtain the actual measured values ​​of range hood power and exhaust volume on each floor in the public flue; the cause-effect analysis module 102 is used to establish a causal effect model between range hood power and exhaust volume according to the actual measured values ​​of range hood power and exhaust volume; the calculation module 103 is used to determine the range hood to be optimized on the target floor according to the causal effect model; the optimization module 104 is used to establish an exhaust volume optimization model based on the number of range hoods to be optimized, and determine the target power of the range hood to be optimized according to the exhaust volume optimization model.

[0129] Optionally, the causal analysis module 102 is configured to: use the actual measured values ​​of the range hood power of each floor except the characteristic floor as input parameters, and the predicted values ​​of the range hood power of the characteristic floor as output parameters to establish a processing variable prediction model; use the actual measured values ​​of the range hood power of all floors as input parameters, and the predicted values ​​of the exhaust volume of the target floor as output parameters to establish a result variable prediction model; and establish a causal effect model based on the processing variable prediction model and the result variable prediction model.

[0130] Optionally, the causal analysis module 102 is also configured to: calculate the residual of the processing variable based on the actual measured value of the range hood power of the feature floor and the predicted value of the processing variable prediction model; calculate the residual of the result variable based on the actual measured value of the exhaust volume of the target floor and the predicted value of the result variable prediction model; establish a causal effect model based on the residual of the processing variable and the residual of the result variable.

[0131] Optionally, the causal analysis module 102 is further configured to: establish a training set and a test set based on the measured values ​​of the range hood power and the exhaust volume; and cross-validate the causal effect model based on the training set and the test set until the causal effect model meets the prediction performance conditions.

[0132] Optionally, the optimization module 104 is configured to: use the range hood power of the range hood to be optimized as the optimization parameter, and the predicted value of the exhaust volume of the target floor as the optimization result, to establish an exhaust volume optimization model; and / or use the actual measured value of the range hood power of each floor except the range hood to be optimized as a fixed parameter, use the range hood power of the range hood to be optimized as the optimization parameter, and the predicted value of the exhaust volume of the target floor as the optimization result, to establish an exhaust volume optimization model.

[0133] Optionally, the data acquisition module 101 is configured to: perform at least one of the following data processing strategies on the measured value of the range hood power and the measured value of the exhaust volume: data cleaning, data denoising and data standardization processing.

[0134] Optionally, the cause-effect analysis module 102 is further configured to: obtain environmental parameters of each floor in the public flue, and optimize the cause-effect effect model according to the environmental parameters.

[0135] Optionally, the causal analysis module 102 is further configured to: use the measured power values ​​and environmental parameters of the range hoods on each floor except the characteristic floor as input parameters, and the predicted power values ​​of the range hoods on the characteristic floor as output parameters, to establish an optimized processing variable prediction model; use the measured power values ​​and environmental parameters of the range hoods on all floors as input parameters, and the predicted exhaust volume values ​​of the target floor as output parameters, to establish an optimized result variable prediction model; and establish a causal effect model based on the optimized processing variable prediction model and the result variable prediction model.

[0136] According to another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned common flue exhaust control method.

[0137] Figure 8 A schematic diagram of the structure of an electronic device for implementing the public flue exhaust control method of an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0138] like Figure 8 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0139] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0140] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the above-mentioned common flue exhaust control method.

[0141] In some embodiments, the above-mentioned common flue exhaust control method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the common flue exhaust control method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the above-mentioned common flue exhaust control method in any other appropriate manner (for example, by means of firmware).

[0142] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0143] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0144] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0145] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0146] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0147] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0148] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0149] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling exhaust of a public flue, characterized in that: include: Obtaining the actual measured values ​​of the range hood power and exhaust volume on each floor in the public flue; Establishing a causal effect model between the range hood power and the exhaust volume according to the measured value of the range hood power and the measured value of the exhaust volume; Determine the range hood to be optimized associated with the target floor according to the causal effect model; An exhaust volume optimization model is established based on the number of the range hoods to be optimized, and the target power of the range hoods to be optimized is determined according to the exhaust volume optimization model.

2. The public flue exhaust control method according to claim 1, characterized in that: The establishing of a causal effect model between the range hood power and the exhaust volume according to the range hood power actual measured value and the exhaust volume actual measured value comprises: Taking the measured values ​​of the range hood power of each floor other than the characteristic floor as input parameters and the predicted values ​​of the range hood power of the characteristic floor as output parameters, a processing variable prediction model is established; The measured values ​​of the range hood power of all floors are used as input parameters, and the predicted value of the exhaust volume of the target floor is used as an output parameter to establish a result variable prediction model; The causal effect model is established based on the treatment variable prediction model and the outcome variable prediction model.

3. The public flue exhaust control method according to claim 2, characterized in that: The establishing of the causal effect model based on the treatment variable prediction model and the outcome variable prediction model comprises: Calculating a processing variable residual based on the measured value of the range hood power on the characteristic floor and the predicted value of the processing variable prediction model; Calculating a result variable residual based on the actual measured value of the exhaust volume of the target floor and the predicted value of the result variable prediction model; The causal effect model is established based on the treatment variable residuals and the outcome variable residuals.

4. The public flue exhaust control method according to claim 1, characterized in that: The establishing of a causal effect model between the range hood power and the exhaust volume according to the range hood power actual measured value and the exhaust volume actual measured value comprises: Establishing a training set and a test set based on the measured value of the range hood power and the measured value of the exhaust volume; The causal effect model is cross-validated based on the training set and the test set until the causal effect model meets the prediction performance condition.

5. The public flue exhaust control method according to claim 1, characterized in that: The step of establishing an exhaust volume optimization model based on the number of the range hoods to be optimized, and determining the target power of the range hoods to be optimized according to the exhaust volume optimization model includes: Taking the range hood power of the range hood to be optimized as the optimization parameter and the exhaust volume prediction value of the target floor as the optimization result, establishing the exhaust volume optimization model; and / or, The exhaust volume optimization model is established by taking the measured values ​​of the range hood power of each floor except the range hood to be optimized as a fixed parameter, taking the range hood power of the range hood to be optimized as an optimization parameter, and taking the predicted value of the exhaust volume of the target floor as the optimization result.

6. The public flue exhaust control method according to claim 1, characterized in that: The obtaining of the actual measured values ​​of the range hood power and exhaust volume on each floor in the public flue comprises: At least one of the following data processing strategies is performed on the measured value of the range hood power and the measured value of the exhaust volume: data cleaning, data denoising and data standardization.

7. The public flue exhaust control method according to any one of claims 1 to 6, characterized in that: Also includes: Obtaining environmental parameters of each floor in the public flue; The causal effect model is optimized according to the environmental parameters.

8. The public flue exhaust control method according to claim 7, characterized in that: The optimizing the causal effect model according to the environmental parameters includes: Taking the measured values ​​of range hood power and environmental parameters of each floor other than the characteristic floor as input parameters, taking the predicted value of range hood power of the characteristic floor as output parameters, and establishing an optimized processing variable prediction model; The measured values ​​of range hood power and environmental parameters of all floors are used as input parameters, and the predicted value of exhaust volume of the target floor is used as output parameter to establish an optimized result variable prediction model; The causal effect model is established based on the optimized treatment variable prediction model and the outcome variable prediction model.

9. A public flue exhaust control device, characterized in that: include: A data acquisition module, used to obtain the actual measured values ​​of the range hood power and exhaust volume on each floor in the public flue; A causal analysis module, used to establish a causal effect model between the range hood power and the exhaust volume according to the measured value of the range hood power and the measured value of the exhaust volume; A calculation module, used for determining the range hood to be optimized associated with the target floor according to the causal effect model; The optimization module is used to establish an exhaust volume optimization model based on the number of the range hoods to be optimized, and determine the target power of the range hoods to be optimized according to the exhaust volume optimization model.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the common flue exhaust control method according to any one of claims 1 to 8.

Images