Smart home system control method and apparatus, and electronic device and smart home system

Abstract

The application discloses a smart home system control method and device, electronic equipment and a smart home system, and belongs to the field of air conditioner control. First, a PMV calculation formula is constructed based on target parameters, and a PMV model is trained, then a first PMV value is calculated according to the PMV calculation formula, and a second PMV value is obtained according to the PMV model; a target PMV value is obtained by comprehensively considering the first PMV value and the second PMV value, and devices in the smart home system are controlled based on the target PMV value and actual target parameters. The application adopts two ways to obtain the target PMV value, so as to solve the problem that formula calculation depends on fixed parameter setting, and the calculated PMV does not match the actual situation, and the obtained target PMV value is accurate, and the control effect is better.

Description

Technical Field

[0001] This application relates to the field of air conditioning control technology, and in particular, to a smart home system control method, device, electronic device, and smart home system. Background Technology

[0002] PMV (Predicted Mean Vote) is a comprehensive evaluation index that takes into account various factors related to human thermal comfort, based on the fundamental equation of human thermal balance and the levels of subjective thermal sensation in psychophysiology. The PMV index represents the average index of a group's votes on seven levels of thermal sensation (from -3 to +3).

[0003] The PMV index is derived by introducing the human body heat load (TL), which reflects the degree of deviation from the body's thermal balance. Its theoretical basis is that when the human body is in a steady-state thermal environment, the greater the heat load, the further the body deviates from a state of thermal comfort. In other words, the larger the positive value of the human body heat load, the hotter the person feels; the larger the negative value, the colder the person feels.

[0004] In some existing air conditioning systems, PMV (partial thermal velocity) is considered in the design, and the PMV value is calculated using a preset formula. The air conditioner is then controlled based on the PMV value to improve the user's thermal comfort. However, these formulas rely on fixed parameter settings, resulting in the calculated PMV not matching the actual situation. Consequently, the control effect of the air conditioner is poor. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this application provides a smart home system control method, device, electronic device, and smart home system. This addresses the problem that some existing air conditioning systems are designed with PMV (Potential Thermal Value) in mind, calculating the PMV value using preset formulas and controlling the air conditioner based on the PMV value to improve user thermal comfort. However, these formulas rely on fixed parameter settings, leading to PMV values ​​that do not match actual conditions, resulting in poor control performance when controlling the air conditioner.

[0006] The technical solution adopted by this application to solve its technical problem is:

[0007] Firstly, a method for controlling a smart home system is provided, including:

[0008] The PMV calculation formula is constructed based on the target parameters and the PMV model is trained. The target parameters include parameters in three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

[0009] Obtain the actual target parameters, and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value;

[0010] The target PMV value is obtained based on the first PMV value and the second PMV value;

[0011] The devices within the smart home system are controlled based on the target PMV value and the actual target parameters.

[0012] Furthermore, the PMV model includes an input layer, a hidden layer, and an output layer;

[0013] The input layer includes n dimensions, where n = p * q, where p is the number of data categories of the actual target parameter based on time, and q is the number of sub-dimensions of the actual target parameter;

[0014] The hidden layers consist of three fully connected layers with 128, 64, and 32 neurons respectively. The activation function for the hidden layers is ReLU, the loss function is cross-entropy loss, and the optimizer is Adam optimizer.

[0015] The softmax function is used to convert the activation values ​​of neurons in the output layer into a probability distribution.

[0016] Furthermore, the PMV calculation formula is as follows:

[0017] PMV1=[a*exp(-b*M)+c]{MWd[ef(MW)-Pa]-h*(MWg)-i*M(h-Pa)-j*M(v-ta)-k*fcl*[(tcl+m)-(+m)]-fcl*hc(tcl-ta)};

[0018] Wherein, PMV1 is the first PMV value, a, b, d, f, h, i, j, k are preset coefficients; c, e, g, m and v are preset values, M is the human body's energy metabolism rate, W is the mechanical work done by the human body, ta is the indoor air temperature, rh is the relative humidity, Pa is the atmospheric pressure, hc is the convective heat transfer coefficient, tcl is the radiative temperature, and fcl is the thermal resistance of clothing.

[0019] Further, obtaining the target PMV value based on the first PMV value and the second PMV value includes:

[0020] Target PMV value = First weight * First PMV value + Second weight * Second PMV value.

[0021] Furthermore, it also includes:

[0022] Obtain the evaluation metrics for the PMV model, wherein the evaluation metrics include at least one of the following: accuracy, precision, and F1 score;

[0023] The value of the second weight is determined based on the evaluation index.

[0024] Furthermore, controlling the devices within the smart home system based on the target PMV value and the actual target parameters includes:

[0025] When the absolute value of the target PMV value is greater than the preset PMV value, the indoor enthalpy value and the outdoor enthalpy value are obtained;

[0026] When the indoor and outdoor enthalpy values ​​do not meet the preset requirements, the windows are closed, and the equipment is controlled according to the actual target parameters; when the indoor and outdoor enthalpy values ​​meet the preset requirements, the windows are opened for continuous ventilation for a preset duration.

[0027] Furthermore, controlling the device according to the actual target parameters includes:

[0028] When the indoor temperature is not equal to the set temperature:

[0029] If the target PMV value is greater than 0, the air outlet temperature of the air conditioner is reduced at a first preset rate until the indoor temperature equals the set temperature.

[0030] If the target PMV value is less than 0, the air outlet temperature of the air conditioner is increased at a second preset rate until the indoor temperature equals the set temperature.

[0031] Furthermore, it also includes:

[0032] First preset rate = First preset coefficient * (Target PMV value - Preset PMV value);

[0033] And / or, the second preset rate = the second preset coefficient * (-target PMV value).

[0034] Furthermore, controlling the device according to the actual target parameters includes:

[0035] When the indoor temperature equals the set temperature:

[0036] If the indoor air velocity is greater than the preset velocity, the air conditioner's output velocity will be reduced to the third preset rate.

[0037] And / or, if the indoor humidity is less than the first preset humidity, the humidifier is controlled to continuously humidify for a preset humidification duration; if the indoor humidity is greater than or equal to the first preset humidity and less than or equal to the second preset humidity, the current humidity is maintained; if the indoor temperature is greater than the second preset humidity, the dehumidifier is controlled to continuously dehumidify for a preset dehumidification duration.

[0038] Furthermore, it also includes:

[0039] Preset humidification duration = Preset humidification coefficient * (First preset humidity - Indoor humidity);

[0040] And / or, preset dehumidification duration = preset dehumidification coefficient * (indoor humidity - second preset humidity).

[0041] Furthermore, it also includes:

[0042] Obtain the absolute value of the difference between the indoor enthalpy value and the outdoor enthalpy value;

[0043] The product of the absolute value and the preset ventilation coefficient is used as the preset ventilation duration.

[0044] Furthermore, it also includes:

[0045] When the target PMV value is greater than 0, if the outdoor enthalpy value is greater than the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the outdoor enthalpy value is less than or equal to the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0046] When the target PMV value is less than 0, if the indoor enthalpy value is greater than the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the indoor enthalpy value is less than or equal to the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0047] Furthermore, controlling the devices within the smart home system based on the target PMV value and the actual target parameters includes:

[0048] When the absolute value of the target PMV value is less than or equal to the preset PMV value, the current state remains unchanged.

[0049] Furthermore, the target parameters include parameters in three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

[0050] Furthermore, when the target parameter is the thermal resistance of clothing, obtaining the actual target parameter includes:

[0051] Obtain information on the current season and the clothing of the target personnel indoors;

[0052] A seasonal coefficient is obtained based on the current seasonal information, and a clothing coefficient is obtained based on the clothing information;

[0053] The thermal resistance of clothing is obtained based on the seasonal coefficient and the clothing coefficient, and the calculation formula is as follows:

[0054] Clothing thermal resistance = seasonal coefficient * clothing coefficient * preset base value.

[0055] Secondly, a smart home system control device is provided, comprising:

[0056] The formula and model module is used to construct the PMV calculation formula and train the PMV model based on the target parameters, which include parameters in three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

[0057] The PMV value acquisition module is used to acquire actual target parameters and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value.

[0058] The target value calculation module is used to obtain a target PMV value based on the first PMV value and the second PMV value;

[0059] The device control module is used to control the devices in the smart home system based on the target PMV value and the actual target parameters.

[0060] Thirdly, an electronic device is provided, comprising:

[0061] At least one processor and at least one memory;

[0062] The memory stores the executable instructions of the processor;

[0063] The processor is configured to perform the smart home system control method described above.

[0064] Fourthly, a smart home system is provided, which applies the aforementioned smart home system control method.

[0065] Beneficial effects:

[0066] This application provides a smart home system control method, device, electronic device, and smart home system. First, a PMV calculation formula is constructed based on target parameters, and a PMV model is trained. Then, a first PMV value is calculated according to the PMV calculation formula, and a second PMV value is obtained according to the PMV model. The first and second PMV values ​​are combined to obtain the target PMV value. Based on the target PMV value and the actual target parameters, the devices within the smart home system are controlled. This application uses two methods to obtain the target PMV value, solving the problem that the formula calculation relies on fixed parameter settings, leading to a discrepancy between the calculated PMV and the actual situation. The obtained target PMV value is accurate, and the control effect is better. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 This is a flowchart of a smart home system control method provided in an embodiment of this application;

[0069] Figure 2 This is a schematic diagram of an initial neural network structure provided in an embodiment of this application;

[0070] Figure 3 This is a schematic diagram of a final PMV model structure provided in an embodiment of this application;

[0071] Figure 4 This is a flowchart illustrating a specific smart home system control method provided in an embodiment of this application;

[0072] Figure 5 This is a schematic diagram of the structure of a smart home system control device provided in an embodiment of this application;

[0073] Figure 6 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0074] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be described in detail below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0075] Reference Figure 1 This application provides a smart home system control method, including:

[0076] S11: Construct the PMV calculation formula and train the PMV model based on the target parameters, which include parameters in three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

[0077] The environmental parameters include, but are not limited to, the following sub-dimensions: indoor temperature, indoor humidity, outdoor temperature, outdoor humidity, wind speed, and atmospheric pressure. These parameters are acquired through sensors located at appropriate locations or via a network.

[0078] Human body parameters include, but are not limited to, the following sub-dimensions: metabolic rate and mechanical work performed by the human body. Metabolic rate is measured by radar, while mechanical work is calculated by acquiring movement trajectories using radar or cameras.

[0079] The PMV model includes an input layer, a hidden layer, and an output layer.

[0080] The input layer comprises n dimensions, where n = p * q, where p is the number of data categories for the actual target parameter based on time, and q is the number of sub-dimensions for the actual target parameter. During data processing, feature analysis is made as comprehensive as possible to ensure the accuracy of the prediction results and the stability of the model operation. The collected data includes the current time t, previous times t-1... and even earlier times tm, totaling p categories, with q data sub-dimensions collected for each category.

[0081] The hidden layers consist of three fully connected layers with 128, 64, and 32 neurons respectively. The ReLU activation function is used, the cross-entropy loss function is used, and the Adam optimizer is used. The initial learning rate is set to 0.002, the number of training epochs is 150, and the batch size is 32. The initial network structure is as follows. Figure 1 As shown.

[0082] The softmax function is used to convert the activation values ​​of neurons in the output layer into a probability distribution. Specifically, as follows... Figure 3 As shown.

[0083] The formula for calculating PMV is as follows:

[0084] PMV1=[a*exp(-b*M)+c]{MWd[ef(MW)-Pa]-h*(MWg)-i*M(h-Pa)-j*M(v-ta)-k*fcl*[(tcl+m)-(+m)]-fcl*hc(tcl-ta)};

[0085] Wherein, PMV1 is the first PMV value, a, b, d, f, h, i, j, k are preset coefficients; c, e, g, m and v are preset values, M is the human body's energy metabolism rate, W is the mechanical work done by the human body, ta is the indoor air temperature, rh is the relative humidity, Pa is the atmospheric pressure, hc is the convective heat transfer coefficient, tcl is the radiative temperature, and fcl is the thermal resistance of clothing.

[0086] S12: Obtain the actual target parameters, and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value;

[0087] In one embodiment, when the target parameter is the thermal resistance of clothing, obtaining the actual target parameter includes:

[0088] Obtain current season information and clothing information of target personnel indoors; obtain a seasonal coefficient based on the current season information and a clothing coefficient based on the clothing information; obtain clothing thermal resistance based on the seasonal coefficient and the clothing coefficient, using the following formula: Clothing thermal resistance = seasonal coefficient * clothing coefficient * preset base value.

[0089] Since the clothing coefficient is obtained through user input or camera acquisition (e.g., the user inputs their own clothing or the camera acquires the clothing the user is wearing, and then the clothing coefficient is obtained based on a preset correspondence), and since the camera's permissions also require user authorization, in practice, the user may not grant authorization or input clothing. Therefore, in another embodiment, when calculating the clothing coefficient, the clothing coefficient can use a default value or not use a clothing coefficient at all.

[0090] In another embodiment, the thermal resistance of clothing can be calculated using only the clothing coefficient.

[0091] S13: Obtain the target PMV value based on the first PMV value and the second PMV value;

[0092] The target PMV value is obtained based on the first PMV value and the second PMV value, including:

[0093] Target PMV value = First weight * First PMV value + Second weight * Second PMV value.

[0094] In one embodiment, the first weight and the second weight are preset fixed values. For example, the first weight = the second weight = 0.5.

[0095] In another embodiment, an evaluation metric for the PMV model is obtained, the evaluation metric including at least one of the following: accuracy, precision, and F1 score; the value of the second weight is determined based on the evaluation metric. This is because the quality of the trained PMV model is uncertain in practice. When the quality is good, the second weight can be larger, i.e., the second weight is greater than the first weight. When the quality is low, the second weight can be smaller, i.e., the second weight is less than the first weight. The specific value of the second weight is determined according to a preset correspondence with the evaluation metric. After determining the second weight, the first weight is then determined. First weight = 1 - second weight.

[0096] S14: Control the devices in the smart home system based on the target PMV value and the actual target parameters.

[0097] The control of devices within the smart home system based on the target PMV value and the actual target parameters includes:

[0098] When the absolute value of the target PMV is greater than the preset PMV, the indoor and outdoor enthalpy values ​​are obtained; when the indoor and outdoor enthalpy values ​​do not meet the preset requirements, the windows are closed, and the device is controlled according to the actual target parameters; preferably, when the indoor temperature is not equal to the set temperature:

[0099] If the target PMV value is greater than 0, the air outlet temperature of the air conditioner is reduced at a first preset rate until the indoor temperature equals the set temperature; if the target PMV value is less than 0, the air outlet temperature of the air conditioner is increased at a second preset rate until the indoor temperature equals the set temperature.

[0100] The first preset rate and the second preset rate can be preset fixed values, or they can be calculated based on the target PMV value. For example, the first preset rate = first preset coefficient * (target PMV value - preset PMV value); and / or, the second preset rate = second preset coefficient * (-target PMV value).

[0101] When the indoor temperature equals the set temperature:

[0102] If the indoor air velocity is greater than the preset velocity, the air conditioner's output velocity is reduced by a third preset rate; and / or, if the indoor humidity is less than the first preset humidity, the humidifier is controlled to continuously humidify for a preset humidification duration; if the indoor humidity is greater than or equal to the first preset humidity and less than or equal to the second preset humidity, the current humidity is maintained; if the indoor temperature is greater than the second preset humidity, the dehumidifier is controlled to continuously dehumidify for a preset dehumidification duration.

[0103] In one embodiment, the preset humidification duration and preset dehumidification duration are preset fixed values. In another embodiment, the preset humidification duration = preset humidification coefficient * (first preset humidity - indoor humidity) and / or, the preset dehumidification duration = preset dehumidification coefficient * (indoor humidity - second preset humidity).

[0104] When the indoor enthalpy and the outdoor enthalpy meet the preset requirements, the window is opened for continuous ventilation for a preset duration.

[0105] In one embodiment, the preset ventilation duration is a preset fixed value. In another embodiment, the absolute value of the difference between the indoor enthalpy and the outdoor enthalpy is obtained; the product of the absolute value and a preset ventilation coefficient is used as the preset ventilation duration.

[0106] Also includes:

[0107] When the target PMV value is greater than 0, if the outdoor enthalpy value is greater than the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the outdoor enthalpy value is less than or equal to the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0108] When the target PMV value is less than 0, if the indoor enthalpy value is greater than the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the indoor enthalpy value is less than or equal to the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0109] The step of controlling the devices within the smart home system based on the target PMV value and the actual target parameters includes:

[0110] When the absolute value of the target PMV value is less than or equal to the preset PMV value, the current state remains unchanged.

[0111] To more clearly illustrate the solution of this application, this application also provides a specific implementation method, including the following steps:

[0112] 1. Data Processing

[0113] 1.1 Data Preparation

[0114] Indoor and outdoor sensor parameters: indoor temperature, indoor humidity, outdoor temperature, outdoor humidity;

[0115] WiFi can be used to obtain location, geographic information, weather information, and seasonal information.

[0116] An anemometer / wind sensor acquires airflow velocity;

[0117] The preferred method is to connect to the internet via Wi-Fi to obtain seasonal information and calculate the thermal resistance of the clothing based on this information. Alternatively, the thermal resistance of the clothing can be obtained by using an RGB camera combined with a target detection algorithm (optional; to protect user privacy, the camera requires user authorization to use, and users can choose to turn off the camera and manually input the relevant thermal resistance parameters). Or, the thermal resistance of the clothing can be determined based on both seasonal information and clothing information obtained by the camera.

[0118] Radar is used to obtain information on the level of human activity and the number / density of people (generally UWB radar or millimeter-wave radar).

[0119] Infrared sensors acquire the temperature of the human body / room area;

[0120] PMV value (calculated using a formula combined with a neural network classification model).

[0121] Data collection can be performed according to Table 1:

[0122] Table 1

[0123] parameter Indoor temperature Indoor humidity outdoor temperature Outdoor humidity Human energy metabolism rate Clothing thermal resistance air wind speed ... PMV describe Indoor temperature Relative humidity of indoor environment Outdoor ambient temperature Relative humidity of outdoor environment Human energy metabolism rate thermal resistance of clothing air velocity ... Predicted average evaluation index unit Celsius percentage Celsius percentage <![CDATA[W / m 2 ]]> Celsius / square meter / hour meters per second ... Unitless, output a specific numerical value.

[0124] 1.2 Method for Calculating the Thermal Resistance of Clothing

[0125] Clothing thermal resistance (rcl) is an indicator of clothing's insulation performance, used to assess its ability to impede heat flow when worn. The following is a calculation method for calculating clothing thermal resistance based on seasonal and clothing information, without a camera or when camera image capture is not permitted. Clothing thermal resistance calculation formula:

[0126] fcl = S*T_season

[0127] Where: S is the seasonal coefficient, which is adjusted according to different seasons, and the values ​​are as follows: Winter: S = 1.5; Spring: S = 1.0; Summer: S = 0.5; Autumn: S = 1.0.

[0128] S2 is the clothing coefficient, which is adjusted according to different clothing types, and its values ​​are as follows: Down jacket:

[0129] T_season is a seasonal temperature index, set based on historical climate data, with the following values:

[0130] Winter: T_season = 20 (assuming average winter temperatures are low)

[0131] Spring: T_season = 10 (Spring temperatures are moderate)

[0132] Summer: T_season = 5 (Summer temperatures are higher)

[0133] Autumn: T_season = 10 (Autumn temperatures are moderate).

[0134] Calculation Example: Assuming it is currently winter, according to the above formula and coefficients, the thermal resistance of the clothing is calculated as follows: fcl = 1.5 * T_season. Substituting the winter value of T_season into the formula: fcl = 1.5 * 20 = 30. This means that in winter, according to our formula, the thermal resistance of the clothing is 30.

[0135] 1.3 Calculation Method of Human Energy Metabolic Rate

[0136] Let MetRate be the human body's metabolic rate (unit: W / m³). 2 REI stands for Radar Energy Index. The calculation formula is as follows:

[0137] MetRate = a1 * REI*b1

[0138] Where: a1 is the proportional coefficient, which we set to 0.1, and b1 is the exponential coefficient, which we set to 1.5.

[0139] Radar Energy Index (REI) range: 0 to 100 is the normal monitoring range.

[0140] Calculation example: Assuming the radar detects a value of REI = 50, the human body's metabolic rate can be calculated using the formula above: MetRate = 0.1 * 50 * 1.5. This means that the human body's metabolic rate can be estimated based on the value detected by the radar.

[0141] 1.4 PMV Calculation Method

[0142] PMV (Predicted Mean Vote) is a comprehensive evaluation index that takes into account many factors related to human thermal comfort, based on the fundamental equation of human thermal balance and the level of subjective thermal sensation in psychophysiology.

[0143] PMV1=[0.303exp(-0.036M)+0.0275]{MW-3.05[5.733-0.007(MW)-Pa]-0.42

[0144] (MW-58.2)-0.0173M(5.867-Pa)-0.0014M(34-ta)-

[0145] 3.9610fcl[(tcl+273)-(+273)]-fcl*hc(tcl-ta)}

[0146] In the formula, M represents the human body's energy metabolism rate, determined by the amount of human activity; W represents the mechanical work performed by the human body; both can be obtained from radar detection and side calculations. Indoor temperature, indoor humidity, outdoor temperature, and outdoor humidity: these parameters will be used as ta (indoor air temperature) and rh (relative humidity) in the PMV formula. WiFi can obtain location, geographic information, weather information, and seasonal information to derive atmospheric pressure Pa and convective heat transfer coefficient hc. Infrared sensors determine the environmental conditions around the human body, including radiation temperature tcl and clothing thermal resistance fcl. Clothing thermal resistance is usually determined by the clothing material and thickness, and can be obtained by manual input by the user or by scanning the clothing label with a camera.

[0147] The thermal sensation corresponding to the PMV index is shown in Table 2 below:

[0148] Table 2

[0149] heat sensation hot warm Slightly warm Moderate Slightly cool cold cold PMV value 3 2 1 0 -1 -2 -3

[0150] In subsequent control, PMV = (model-predicted PMV1 + formula-calculated PMV2) / 2.

[0151] 1.5 Enthalpy Calculation

[0152] Enthalpy in air generally refers to the total heat content of air. Enthalpy is a measure of the combined heat and humidity of a unit of air, usually represented by the symbol 'i' and measured in kJ / kg; air temperature is represented by the symbol 't' and measured in °C; and the humidity content of the air is represented by the symbol 'd1' and measured in kg. The formula for calculating enthalpy is as follows:

[0153] i = 1.01t + (2490 + 1.84t) * d1

[0154] 1.6 Calculation methods for other indicators

[0155] Other indicators can be calculated in the same way as above. For those that cannot be calculated or are difficult to calculate in practice, specific default parameter values ​​can be set.

[0156] 2 Machine Learning Classification Algorithms

[0157] PMV is divided into two types: calculated PMV (i.e., the first PMV value) and model-predicted PMV (i.e., the second PMV value). The average of these two methods is used to improve the stability of PMV calculation and prediction. However, the specific PMV calculation formula described above is often difficult to utilize in practice due to cost and other issues, and default parameters are mostly used. Therefore, a neural network classification algorithm is an alternative PMV calculation method. The corrected PMV value is obtained by averaging the formula calculation result and the algorithm classification result.

[0158] During data processing, feature analysis is conducted as comprehensively as possible to ensure the accuracy of prediction results and the stability of model operation. The collected data includes the current time t, previous times t-1, ..., even earlier times tm, totaling p categories, with q data dimensions collected for each category, as shown in Table 3.

[0159] Table 3

[0160]

[0161] The total input dimension is p*q=n, and the output is a one-dimensional PMV. A multilayer perceptron mechanism is trained using the TensorFlow machine learning training platform. The input layer contains n inputs, set according to specific needs; the output layer contains 4 neurons, and the softmax function is used to convert the activation values ​​of the output layer neurons into a probability distribution. There are 3 hidden layers with 128, 64, and 32 neurons respectively. The ReLU activation function is used for the hidden layers, the cross-entropy loss function is used, and the Adam optimizer is used. The initial learning rate is set to 0.002, the number of training epochs is 150, and the batch size is 32. The specific network structure is as follows: Figure 2 As shown.

[0162] The middle hidden layer is a three-layer fully connected layer, using the ReLU activation function. However, in the PMV prediction task, the goal is to predict values ​​for seven specific categories ranging from -3 to 3, which is a classification problem. Therefore, the network structure can be... Figure 3 As shown, the specific structure is as follows:

[0163] Input layer: The input layer receives the aforementioned raw data.

[0164] Fully connected layers: From the input end to the output end, the three fully connected layers have 128, 64, and 32 neurons respectively.

[0165] Output layer: Finally, depending on the type of prediction task, a Softmax output layer is used to generate the final prediction result.

[0166] In this task, the model aims to learn the relationship between environmental parameters and input features PMV, and to optimize the network weights using training data so as to accurately predict PMV parameters.

[0167] Control logic diagram as follows Figure 4 As shown:

[0168] (1) Initial settings

[0169] Initial indoor temperature setting: Set to 26℃ to ensure a comfortable environment.

[0170] (2) PMV value detection and adjustment

[0171] If the indoor temperature is not equal to 26℃ and the PMV > 0.5: Adjust the outlet temperature to increase the temperature and reduce the PMV value. Formula: ΔT = 0.1(PMV - 0.5)℃.

[0172] If -0.5≤PMV≤0.5: Maintain the current outlet temperature and do not make any changes.

[0173] If PMV < -0.5: Lower the outlet temperature to increase the PMV value. Formula: ΔT = 0.1(-PMV)℃.

[0174] (3) Wind speed monitoring and adjustment

[0175] When the indoor temperature is 26℃, if V > 0.8 m / s: gradually reduce the wind speed until V ≤ 0.8 m / s. Adjustment steps: V_new = V - 0.1 m / s.

[0176] (4) Indoor humidity adjustment

[0177] When the indoor temperature is 26℃, if the humidity inside Phi is less than 45%, turn on the humidifier to replenish the humidity. Estimated humidification duration: t_humidify = 60(45 - inside Phi) seconds.

[0178] If 45% ≤ Phi ≤ 65%: Maintain the current humidity level and do not take any action.

[0179] If the humidity inside Phi is >65%, activate the dehumidifier to reduce humidity. Estimated humidification duration: t_dehumidify = 60(Phi - 65) seconds.

[0180] (5) Ventilation logic

[0181] Indoor and outdoor humidity comparison and ventilation: Judgment condition: If ioutdoor > iindoor and PMV > 0, then open the window for ventilation. Ventilation time estimation: t_ventilate = 30|iindoor - ioutdoor| minutes.

[0182] The comprehensive steps are as follows: Check PMV, wind speed, and humidity, and gradually adjust each parameter. Through multi-dimensional control and regulation, the dimensions that mainly affect PMV values ​​are adjusted to ensure indoor comfort. Settings are dynamically updated based on real-time data to continuously optimize the indoor environment. This logical design ensures automatic adjustment and optimization of the indoor environment under different conditions, improving living comfort and air quality.

[0183] The innovation of this application lies in its comprehensive application of advanced sensor technology, machine learning algorithms, and dynamic control logic to achieve more efficient and personalized environmental control. First, by integrating radar monitoring and dynamic calculation of clothing thermal resistance, this proposal can more accurately assess human thermal comfort, a key factor in improving system performance. Second, the application of neural networks to optimize PMV improves prediction accuracy and the level of intelligence in environmental control. Finally, the implementation of dynamic environmental control logic based on real-time data ensures that the indoor environment can be adjusted instantly according to real-time monitoring data, thereby maximizing user comfort and energy efficiency.

[0184] The advantages of this application are as follows: 1) Dynamic clothing thermal resistance calculation: By using seasonal information and RGB camera data or user input, the thermal resistance of clothing is adjusted in real time, improving the accuracy of thermal comfort assessment; 2) Comprehensive environmental monitoring: Integrating multiple sensors such as indoor and outdoor temperature, humidity, and wind speed, providing more detailed environmental data; 3) Accurate PMV prediction: Utilizing neural networks to optimize PMV calculation, enhancing the accuracy and adaptability of predictions; 4) Intelligent control logic: Dynamically adjusting environmental settings based on real-time data, achieving more flexible environmental control. These advantages all stem from the adoption of advanced sensor integration, machine learning algorithms, and intelligent control strategies.

[0185] Based on the same inventive concept, such as Figure 5 As shown in the figure, this application embodiment also provides a smart home system control device 50, including:

[0186] Formula and model module 51 is used to construct the PMV calculation formula based on the target parameters and train the PMV model. The target parameters include parameters in three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

[0187] The PMV model includes an input layer, a hidden layer, and an output layer.

[0188] The input layer includes n dimensions, where n = p * q, where p is the number of data categories of the actual target parameter based on time, and q is the number of sub-dimensions of the actual target parameter;

[0189] The hidden layers consist of three fully connected layers with 128, 64, and 32 neurons respectively. The activation function for the hidden layers is ReLU, the loss function is cross-entropy loss, and the optimizer is Adam optimizer.

[0190] The softmax function is used to convert the activation values ​​of neurons in the output layer into a probability distribution.

[0191] The formula for calculating PMV is as follows:

[0192] PMV1=[a*exp(-b*M)+c]{MWd[ef(MW)-Pa]-h*(MWg)-i*M(h-Pa)-j*M(v-ta)-k*fcl*[(tcl+m)-(+m)]-fcl*hc(tcl-ta)};

[0193] Wherein, PMV1 is the first PMV value, a, b, d, f, h, i, j, k are preset coefficients; c, e, g, m and v are preset values, M is the human body's energy metabolism rate, W is the mechanical work done by the human body, ta is the indoor air temperature, rh is the relative humidity, Pa is the atmospheric pressure, hc is the convective heat transfer coefficient, tcl is the radiative temperature, and fcl is the thermal resistance of clothing.

[0194] The PMV value acquisition module 52 is used to acquire actual target parameters and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value.

[0195] When the target parameter is the thermal resistance of clothing, obtaining the actual target parameter includes:

[0196] Obtain information on the current season and the clothing of the target personnel indoors;

[0197] A seasonal coefficient is obtained based on the current seasonal information, and a clothing coefficient is obtained based on the clothing information;

[0198] The thermal resistance of clothing is obtained based on the seasonal coefficient and the clothing coefficient, and the calculation formula is as follows:

[0199] Clothing thermal resistance = seasonal coefficient * clothing coefficient * preset base value.

[0200] In practical use, the preset base value can be determined based on either the current season information or clothing information. Alternatively, the user can directly input the value. The preset base value can be the same for all four seasons, or different values ​​can be set for different seasons.

[0201] Target value calculation module 53 is used to obtain a target PMV value based on the first PMV value and the second PMV value;

[0202] The process of obtaining the target PMV value based on the first PMV value and the second PMV value includes:

[0203] Target PMV value = First weight * First PMV value + Second weight * Second PMV value.

[0204] In one embodiment, the first weight and the second weight are preset fixed values.

[0205] In another embodiment, an evaluation metric for the PMV model is obtained, the evaluation metric including at least one of the following: accuracy, precision, and F1 score;

[0206] The value of the second weight is determined based on the evaluation index. First weight = 1 - second weight.

[0207] The device control module 54 is used to control the devices in the smart home system based on the target PMV value and the actual target parameters.

[0208] The control of devices within the smart home system based on the target PMV value and the actual target parameters includes:

[0209] When the absolute value of the target PMV value is greater than the preset PMV value, the indoor enthalpy value and the outdoor enthalpy value are obtained;

[0210] When the indoor and outdoor enthalpy values ​​do not meet the preset requirements, the windows are closed, and the equipment is controlled according to the actual target parameters; when the indoor and outdoor enthalpy values ​​meet the preset requirements, the windows are opened for continuous ventilation for a preset duration.

[0211] Controlling the device based on the actual target parameters includes:

[0212] When the indoor temperature is not equal to the set temperature:

[0213] If the target PMV value is greater than 0, the air outlet temperature of the air conditioner is reduced at a first preset rate until the indoor temperature equals the set temperature.

[0214] If the target PMV value is less than 0, the air outlet temperature of the air conditioner is increased at a second preset rate until the indoor temperature equals the set temperature.

[0215] In one embodiment, the first preset rate and the second preset rate are preset fixed values.

[0216] In another embodiment, it also includes:

[0217] First preset rate = First preset coefficient * (Target PMV value - Preset PMV value);

[0218] And / or, the second preset rate = the second preset coefficient * (-target PMV value).

[0219] Furthermore, controlling the device according to the actual target parameters includes:

[0220] When the indoor temperature equals the set temperature:

[0221] If the indoor air velocity is greater than the preset velocity, the air conditioner's output velocity will be reduced to the third preset rate.

[0222] And / or, if the indoor humidity is less than the first preset humidity, the humidifier is controlled to continuously humidify for a preset humidification duration; if the indoor humidity is greater than or equal to the first preset humidity and less than or equal to the second preset humidity, the current humidity is maintained; if the indoor temperature is greater than the second preset humidity, the dehumidifier is controlled to continuously dehumidify for a preset dehumidification duration.

[0223] In one embodiment, the preset humidification duration and the preset dehumidification duration are preset fixed values.

[0224] In another embodiment, the preset humidification duration = preset humidification coefficient * (first preset humidity - indoor humidity);

[0225] And / or, preset dehumidification duration = preset dehumidification coefficient * (indoor humidity - second preset humidity).

[0226] In one embodiment, the preset ventilation duration is a preset fixed value.

[0227] In another embodiment, the absolute value of the difference between the indoor enthalpy and the outdoor enthalpy is obtained;

[0228] The product of the absolute value and the preset ventilation coefficient is used as the preset ventilation duration.

[0229] Also includes:

[0230] When the target PMV value is greater than 0, if the outdoor enthalpy value is greater than the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the outdoor enthalpy value is less than or equal to the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0231] When the target PMV value is less than 0, if the indoor enthalpy value is greater than the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; if the indoor enthalpy value is less than or equal to the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

[0232] When the absolute value of the target PMV value is less than or equal to the preset PMV value, the current state remains unchanged.

[0233] The smart home system control device provided in this application first constructs a PMV calculation formula and trains a PMV model based on target parameters. Then, it calculates a first PMV value according to the PMV calculation formula and a second PMV value according to the PMV model. The first and second PMV values ​​are combined to obtain a target PMV value. Based on the target PMV value and the actual target parameters, the device within the smart home system is controlled. This application's solution uses two methods to obtain the target PMV value, solving the problem that the formula calculation relies on fixed parameter settings, leading to a discrepancy between the calculated PMV and the actual situation. The obtained target PMV value is accurate, resulting in better control performance.

[0234] Based on the same inventive concept, such as Figure 6 As shown, this application embodiment also provides an electronic device 60, including:

[0235] At least one processor 61 and at least one memory 62;

[0236] The memory stores the executable instructions of the processor;

[0237] The processor is configured to execute the smart home system control method provided in the above embodiments.

[0238] The electronic device provided in this application stores executable instructions of a processor in a memory. When these executable instructions are executed, the processor first constructs a PMV calculation formula and trains a PMV model based on target parameters. Then, it calculates a first PMV value according to the PMV calculation formula and a second PMV value according to the PMV model. The first and second PMV values ​​are combined to obtain a target PMV value. Based on the target PMV value and the actual target parameters, the device within the smart home system is controlled. This application's solution uses two methods to obtain the target PMV value, solving the problem that the formula calculation relies on fixed parameter settings, leading to a discrepancy between the calculated PMV and the actual situation. The obtained target PMV value is accurate, resulting in better control.

[0239] Based on the same inventive concept, this application also provides a smart home system that applies the smart home system control method provided in the above embodiments.

[0240] The smart home system provided in this application, by applying the smart home system control method provided in the above embodiments, can first construct a PMV calculation formula and train a PMV model based on target parameters, then calculate a first PMV value according to the PMV calculation formula, and obtain a second PMV value according to the PMV model; combine the first PMV value and the second PMV value to obtain a target PMV value, and control the devices in the smart home system based on the target PMV value and the actual target parameters. This application's solution uses two methods to obtain the target PMV value to solve the problem that the formula calculation relies on fixed parameter settings, leading to a discrepancy between the calculated PMV and the actual situation. The obtained target PMV value is accurate, and the control effect is better.

[0241] It should be noted that in the description of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means at least two.

[0242] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

Claims

1. A method for controlling a smart home system, characterized in that, include: The PMV calculation formula is constructed based on the target parameters, and the PMV model is trained. Obtain the actual target parameters, and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value; The target PMV value is obtained based on the first PMV value and the second PMV value; The devices within the smart home system are controlled based on the target PMV value and the actual target parameters; The control of devices within the smart home system based on the target PMV value and the actual target parameters includes: When the absolute value of the target PMV is greater than the preset PMV, obtain the indoor and outdoor enthalpy values; when the indoor and outdoor enthalpy values ​​do not meet the preset requirements, close the windows, and when the indoor temperature is not equal to the set temperature: If the target PMV value is greater than 0, the air outlet temperature of the air conditioner is reduced at a first preset rate until the indoor temperature equals the set temperature. If the target PMV value is less than 0, the air outlet temperature of the air conditioner is increased at a second preset rate until the indoor temperature equals the set temperature. Wherein, the first preset rate = the first preset coefficient * (target PMV value - preset PMV value); And / or, the second preset rate = the second preset coefficient * (-target PMV value).

2. The method according to claim 1, characterized in that: The PMV model includes an input layer, a hidden layer, and an output layer; The input layer includes n dimensions, where n = p * q, where p is the number of data categories of the actual target parameter based on time, and q is the number of sub-dimensions of the actual target parameter; The hidden layers consist of three fully connected layers with 128, 64, and 32 neurons respectively. The activation function for the hidden layers is ReLU, the loss function is cross-entropy loss, and the optimizer is Adam optimizer. The softmax function is used to convert the activation values ​​of neurons in the output layer into a probability distribution.

3. The method according to claim 1, characterized in that: The formula for calculating PMV is as follows: PMV1=[a*exp(-b*M)+c]{MWd[ef(MW)-Pa]-h*(MWg)-i*M(h-Pa)-j*M(v-ta)-k*fcl*[(tcl+m)]-fcl*hc(tcl-ta)}; Wherein, PMV1 is the first PMV value, a, b, d, f, h, i, j, k are preset coefficients; c, e, g, m and v are preset values, M is the human body's energy metabolism rate, W is the mechanical work done by the human body, ta is the indoor air temperature, rh is the relative humidity, Pa is the atmospheric pressure, hc is the convective heat transfer coefficient, tcl is the radiative temperature, and fcl is the thermal resistance of clothing.

4. The method according to claim 1, characterized in that: The process of obtaining the target PMV value based on the first PMV value and the second PMV value includes: Target PMV value = First weight * First PMV value + Second weight * Second PMV value.

5. The method according to claim 4, characterized in that, Also includes: Obtain the evaluation metrics for the PMV model, wherein the evaluation metrics include at least one of the following: accuracy, precision, and F1 score; The value of the second weight is determined based on the evaluation index.

6. The method according to claim 1, characterized in that: The control of devices within the smart home system based on the target PMV value and the actual target parameters includes: When the indoor enthalpy and the outdoor enthalpy meet the preset requirements, the window is opened for continuous ventilation for a preset duration.

7. The method according to claim 6, characterized in that: Controlling the device according to the actual target parameters includes: When the indoor temperature equals the set temperature: If the indoor air velocity is greater than the preset velocity, the air conditioner's output velocity will be reduced to the third preset rate. And / or, if the indoor humidity is less than the first preset humidity, the humidifier is controlled to continuously humidify for a preset humidification duration; if the indoor humidity is greater than or equal to the first preset humidity and less than or equal to the second preset humidity, the current humidity is maintained; if the indoor humidity is greater than the second preset humidity, the dehumidifier is controlled to continuously dehumidify for a preset dehumidification duration.

8. The method according to claim 7, characterized in that, Also includes: Preset humidification duration = Preset humidification coefficient * (First preset humidity - Indoor humidity); And / or, preset dehumidification duration = preset dehumidification coefficient * (indoor humidity - second preset humidity).

9. The method according to claim 6, characterized in that, Also includes: Obtain the absolute value of the difference between the indoor enthalpy value and the outdoor enthalpy value; The product of the absolute value and the preset ventilation coefficient is used as the preset ventilation duration.

10. The method according to claim 6, characterized in that, Also includes: When the target PMV value is greater than 0, if the outdoor enthalpy value is greater than the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; If the outdoor enthalpy value is less than or equal to the indoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements; When the target PMV value is less than 0, if the indoor enthalpy value is greater than the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value meet the preset requirements; If the indoor enthalpy value is less than or equal to the outdoor enthalpy value, then the indoor enthalpy value and the outdoor enthalpy value do not meet the preset requirements.

11. The method according to claim 1, characterized in that, The control of devices within the smart home system based on the target PMV value and the actual target parameters includes: When the absolute value of the target PMV value is less than or equal to the preset PMV value, the current state remains unchanged.

12. The method according to claim 1, characterized in that, The target parameters include parameters from three dimensions: environmental parameters, human body parameters, and clothing thermal resistance.

13. The method according to claim 12, characterized in that, When the target parameter is the thermal resistance of clothing, obtaining the actual target parameter includes: Obtain information on the current season and the clothing of the target personnel indoors; A seasonal coefficient is obtained based on the current seasonal information, and a clothing coefficient is obtained based on the clothing information; The thermal resistance of clothing is obtained based on the seasonal coefficient and the clothing coefficient, and the calculation formula is as follows: Clothing thermal resistance = seasonal coefficient * clothing coefficient * preset base value.

14. A smart home system control device, characterized in that, include: The formula and model module is used to construct the PMV calculation formula based on the target parameters and to train the PMV model. The PMV value acquisition module is used to acquire actual target parameters and input the actual target parameters into the PMV calculation formula and the PMV model respectively to obtain the first PMV value and the second PMV value. The target value calculation module is used to obtain a target PMV value based on the first PMV value and the second PMV value; The device control module is used to control the devices in the smart home system based on the target PMV value and the actual target parameters; The control of devices within the smart home system based on the target PMV value and the actual target parameters includes: When the absolute value of the target PMV is greater than the preset PMV, obtain the indoor and outdoor enthalpy values; when the indoor and outdoor enthalpy values ​​do not meet the preset requirements, close the windows, and when the indoor temperature is not equal to the set temperature: If the target PMV value is greater than 0, the air outlet temperature of the air conditioner is reduced at a first preset rate until the indoor temperature equals the set temperature. If the target PMV value is less than 0, the air outlet temperature of the air conditioner is increased at a second preset rate until the indoor temperature equals the set temperature. Wherein, the first preset rate = the first preset coefficient * (target PMV value - preset PMV value); And / or, the second preset rate = the second preset coefficient * (-target PMV value).

15. An electronic device, characterized in that, include: At least one processor and at least one memory; The memory stores the executable instructions of the processor; The processor is configured to perform the method according to any one of claims 1-13.

16. A smart home system, characterized in that, The method described in any one of claims 1-13.

Images