Intelligent control method and system for floor radiant air conditioning system

Abstract

The application provides a floor radiation air conditioning system intelligent control method and system, in which indoor personnel load estimation data and thermal comfort calculation data are acquired to estimate the indoor personnel load; a thermal sensation prediction model is established based on indoor environment parameters and the thermal comfort calculation data; an indoor environment and energy consumption prediction model is established based on indoor and outdoor environment parameters, equipment parameters and air conditioning system operation parameters, so that the indoor environment after adjustment of the air conditioning system and the energy consumption of the air conditioning system are predicted; indoor and outdoor environment parameters, indoor personnel load, thermal sensation and air conditioning system energy consumption are used to respectively construct a personnel thermal comfort target function and an energy consumption target function, and a total target function is determined according to the weight of the personnel thermal comfort target function and the energy consumption target function; the improved Melolontha optimization algorithm is applied to solve the total target function, so that the optimal regulation and control parameters of the floor radiation air conditioning system control are obtained, and the reliability of the floor radiation air conditioning system control is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air conditioning system control, in particular to a floor radiation air conditioning system intelligent control method and system. BACKGROUND

[0002] With the rapid development of building intelligent technology, people's demand for building intelligence is getting higher and higher. Intelligent adjustment of air conditioner is an important part of building intelligence, and intelligent air conditioning system can provide healthy and comfortable environment for users to meet their needs in work and life. Most of the current air conditioning systems control the environmental comfort by setting a single temperature and humidity index, which cannot well reflect the individual comfort demand, and is easy to cause "overcooling" and "overheating" phenomenon, resulting in energy waste.

[0003] On the other hand, the traditional floor radiation air conditioning system control strategy takes temperature and dew risk as indicators, ignoring the uncertainty of personnel flow, and the radiation floor has slow cooling supply due to its inherent thermal inertia. When the indoor personnel load changes rapidly, the radiation air conditioning system terminal has a certain lag effect on the adjustment of indoor environment, resulting in a mismatch between supply and demand, and thus leading to poor indoor environmental comfort and energy waste in a certain period of time, so that the reliability of the floor radiation air conditioning system control is not high.

[0004] In order to solve the problem of low reliability of floor radiation air conditioning system control in the prior art, the present application is proposed. SUMMARY

[0005] In order to solve the problems existing in the prior art, the present application provides a floor radiation air conditioning system intelligent control method and system, which effectively solves the problem of low reliability of floor radiation air conditioning system control caused by the prior art, and effectively improves the reliability of floor radiation air conditioning system control.

[0006] The first aspect of the present application provides a floor radiation air conditioning system intelligent control method, comprising:

[0007] Obtaining indoor environment parameters, outdoor environment parameters, thermal imaging video stream, non-thermal imaging video stream and air conditioning system operation data;

[0008] Using computer vision technology to process the thermal imaging video stream and the non-thermal imaging video stream, tracking the indoor personnel, obtaining indoor personnel load estimation data and thermal comfort calculation data, and estimating the indoor personnel load according to the indoor personnel load estimation data;

[0009] Based on the indoor environment parameters and the thermal comfort calculation data, a thermal sensation prediction model is established to predict the indoor personnel thermal sensation;

[0010] Based on indoor and outdoor environment parameters, equipment parameters, air conditioning system operation parameters, an indoor environment and energy consumption prediction model is established, and the indoor environment and air conditioning system energy consumption after adjustment of the air conditioning system are predicted;

[0011] The constraint condition is determined, and a personnel thermal comfort target function and an energy consumption target function are respectively constructed based on the indoor and outdoor environment parameters, the indoor personnel load and the thermal sensation and the air conditioning system energy consumption, and a total target function is determined according to the weights of the personnel thermal comfort target function and the energy consumption target function;

[0012] The improved melittin optimization algorithm is applied to solve the total target function, and the optimal control parameter of the floor radiant air conditioning system control is obtained.

[0013] The second aspect of the present application provides a floor radiant air conditioning system intelligent control system, comprising:

[0014] The first acquisition module acquires indoor and outdoor environment parameters, thermal imaging video stream, non-thermal imaging video stream, air conditioning system operation data and equipment parameters.

[0015] The second acquisition module processes the thermal imaging video stream and the non-thermal imaging video stream using computer vision technology, tracks the indoor personnel, acquires indoor personnel load estimation data and thermal comfort calculation data, and estimates the indoor personnel load according to the indoor personnel load estimation data.

[0016] The first establishment module establishes a thermal sensation prediction model based on the indoor environment parameters and the thermal comfort calculation data, and predicts the indoor personnel thermal sensation.

[0017] The second establishment module establishes an indoor environment and energy consumption prediction model based on the indoor and outdoor environment parameters, the equipment parameters and the air conditioning system operation parameters, and predicts the indoor environment and the air conditioning system energy consumption after adjustment of the air conditioning system.

[0018] The determination module determines the constraint condition, and respectively constructs a personnel thermal comfort target function and an energy consumption target function based on the indoor and outdoor environment parameters, the indoor personnel load, the thermal sensation and the air conditioning system energy consumption, and determines a total target function according to the weights of the personnel thermal comfort target function and the energy consumption target function.

[0019] The solving module applies the improved melittin optimization algorithm to solve the total target function, and obtains the optimal control parameter of the floor radiant air conditioning system control.

[0020] The technical scheme adopted by the present application has the following technical effects:

[0021] 1、The technical scheme of the present application uses computer vision technology to process thermal imaging video streams and non-thermal imaging video streams, track indoor personnel, obtain indoor personnel load estimation data and thermal comfort calculation data, estimates indoor personnel load based on indoor personnel load estimation data, establishes a thermal sensation prediction model based on indoor environmental parameters and thermal comfort calculation data, predicts indoor personnel thermal sensation, establishes an indoor environment and energy consumption prediction model based on indoor and outdoor environmental parameters, equipment parameters and air conditioning system operating parameters, predicts indoor environment and air conditioning system energy consumption after adjustment, determines constraints, constructs personnel thermal comfort objective functions and energy consumption objective functions based on indoor and outdoor environmental parameters, indoor personnel load, thermal sensation and air conditioning system energy consumption, determines a total objective function based on the weights of personnel thermal comfort objective functions and energy consumption objective functions, and obtains optimal control parameters of floor radiant air conditioning system control by solving the total objective function using an improved beetle optimization algorithm, which can adjust radiant air conditioning system parameters in real time according to indoor personnel load and personnel thermal comfort, effectively solves the problem of low reliability of floor radiant air conditioning system control caused by existing technologies, and effectively improves the reliability of floor radiant air conditioning system control.

[0022] 2、The technical scheme of the present application establishes a personnel detection and tracking model based on computer vision technology, obtains the positions and numbers of detected personnel, obtains personnel gender and age range based on human face recognition results, continuously tracks recognized human body parts, recognizes human body actions, obtains unit time movement distance, identifies human body activity state based on human body actions and unit time movement distance of human body parts, estimates indoor personnel load, and can estimate indoor personnel comfort according to personnel type and activity state, thereby further improving the reliability of floor radiant air conditioning system control.

[0023] 3、The technical scheme of the present application can accurately reflect user thermal sensation by real-time monitoring of indoor and outdoor environmental parameters and thermal imaging data, using artificial neural network algorithm, establishing a thermal sensation prediction model based on indoor environmental parameters and thermal comfort calculation data, and predicting indoor personnel thermal sensation.

[0024] 4、The technical scheme of the present application uses computer vision to track and identify indoor personnel numbers and activities, responds to indoor personnel load changes in a timely manner, rapidly adjusts air conditioning parameters, effectively alleviates the hysteresis effect in radiant air conditioning system control, and reduces discomfort and energy waste caused by unbalanced air conditioning load supply and demand.

[0025] 5、The technical scheme of the present application utilizes an artificial neural network algorithm to predict the regulation and control effect of a radiant air conditioning system, and simultaneously determines the constraint conditions, to construct a personnel thermal comfort target function, an energy consumption target function, and a total target function according to the weights of the personnel thermal comfort target function and the energy consumption target function, with the indoor and outdoor environmental parameters, the indoor personnel load, the thermal sensation, and the air conditioning system energy consumption, respectively, and to generate optimal regulation and control parameters by applying an improved Stag Beetle Optimization algorithm, so as to ensure personnel thermal comfort and reduce the energy consumption of the air conditioning system.

[0026] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below, and obviously, other drawings can also be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0028] Figure 1 A flowchart of the method of embodiment one in the present application scheme;

[0029] Figure 2 A flowchart of step S2 in the method of embodiment one in the present application scheme;

[0030] Figure 3 A personnel load diagram corresponding to different activity states of different groups of people in the method of embodiment one in the present application scheme;

[0031] Figure 4 A flowchart of step S3 in the method of embodiment one in the present application scheme;

[0032] Figure 5 A diagram of collecting thermal sensation data in the method of embodiment one in the present application scheme;

[0033] Figure 6 A flowchart of step S4 in the method of embodiment one in the present application scheme;

[0034] Figure 7 A flowchart of step S5 in the method of embodiment one in the present application scheme;

[0035] Figure 8 A flowchart of the Stag Beetle Optimization algorithm in the method of embodiment one in the present application scheme;

[0036] Figure 9 Another flowchart of the method of embodiment one in the present application scheme;

[0037] Figure 10A flowchart for implementing step S7 in the method of example one in the present solution is shown.

[0038] Figure 11 A structural schematic diagram of the system of example two in the present solution is shown. DETAILED DESCRIPTION

[0039] To clearly illustrate the technical features of the present solution, the present solution will be described in detail below through specific implementation manners, and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the present solution. In order to simplify the disclosure of the present solution, the components and settings of specific examples are described below. In addition, reference numerals and / or letters can be repeatedly referred to in different examples. Such repetition is for the purpose of simplification and clarity, and does not in itself indicate a relationship between the various embodiments and / or settings being discussed. It should be noted that the components illustrated in the drawings are not necessarily drawn to scale. The present solution omits the description of well-known components and processing techniques and processes to avoid unnecessarily limiting the present solution.

[0040] Example one

[0041] As shown in Figure 1 , the present solution provides an intelligent control method for a floor radiant air conditioning system, comprising:

[0042] S1, obtaining indoor environment parameters, outdoor environment parameters, thermal imaging video stream, non-thermal imaging video stream, and air conditioning system operation data;

[0043] S2, using computer vision technology to process the thermal imaging video stream and the non-thermal imaging video stream, tracking indoor personnel, obtaining indoor personnel load estimation data and thermal comfort calculation data, and estimating indoor personnel load according to the indoor personnel load estimation data;

[0044] S3, establishing a thermal sensation prediction model based on the indoor environment parameters and the thermal comfort calculation data, and predicting indoor personnel thermal sensation;

[0045] S4, establishing an indoor environment and energy consumption prediction model based on indoor and outdoor environment parameters, equipment parameters, and air conditioning system operation parameters, and predicting the indoor environment after adjustment by the air conditioning system and the energy consumption of the air conditioning system;

[0046] S5, determining constraint conditions, constructing a personnel thermal comfort objective function and an energy consumption objective function based on indoor and outdoor environment parameters, indoor personnel load, thermal sensation, and air conditioning system energy consumption, respectively, and determining a total objective function according to the weights of the personnel thermal comfort objective function and the energy consumption objective function;

[0047] S6, applying an improved scarab beetle optimization algorithm to solve the total objective function, and obtaining optimal control parameters for the floor radiant air conditioning system control.

[0048] Wherein, in step S1, the air conditioning system operation data includes: fresh air temperature, fresh air speed, fresh air volume of fresh air fan, radiant floor surface temperature, radiant floor water supply temperature, radiant floor return water temperature; indoor environmental parameters include: indoor temperature, indoor humidity, wall surface temperature, other surface temperature except wall surface and floor surface; outdoor environmental parameters include: solar radiation, outdoor temperature, outdoor humidity; indoor personnel load estimation data includes the number of personnel, gender, age range, human activity state; thermal comfort calculation data includes human face temperature, human hand temperature, indoor mean radiant temperature, operating temperature, which is the temperature of the combined action of indoor air temperature and indoor mean radiant temperature on human body; indoor and outdoor environmental parameters, thermal imaging video stream, ordinary video stream (non-thermal imaging video stream), air conditioning system operation data are obtained using sensors, and the thermal imaging real-time video is recorded by a thermal imager. The room video is recorded in real time using a thermal imager; the equipment parameters include: fresh air fan speed range, radiant floor water supply temperature range.

[0049] The applied radiant air conditioning system includes: a fresh air fan, a radiant floor, a mixed water pump, a temperature and humidity sensor, a thermal imager, and a camera.

[0050] The obtained data is divided into two parts, one part is historical data, and the historical data is used for training an artificial neural network model; the other part is real-time data, which is used for real-time calculation of the floor radiant air conditioning system intelligent control method and system based on computer vision to solve the optimal control parameters.

[0051] As shown in Figure 2 S2 specifically includes:

[0052] S21, a personnel detection model, a personnel tracking model, and a human body recognition model are established based on computer vision technology, and indoor thermal imaging video stream and non-thermal imaging video stream are input into the trained personnel detection model, personnel tracking model, and human body recognition model;

[0053] S22, the indoor thermal imaging video stream and the non-thermal imaging video stream are normalized, and the trained personnel detection model and personnel tracking model are used for identification to obtain the detected personnel position and the number of personnel;

[0054] The personnel detection model can identify human body in the thermal imaging video stream and the non-thermal imaging video stream, and the personnel tracking model can identify human body movement in the thermal imaging video stream and the non-thermal imaging video stream. Here, it is a neural network based on deep learning (annotated, trained, tested, etc.), and existing neural network models can also be used as long as personnel detection and tracking can be achieved. The present application does not limit this.

[0055] S23, input the personnel position area into the trained human body recognition model, recognize the human body parts, obtain the face temperature according to the temperature values of different positions of the human face, obtain the hand temperature according to the temperature values of different positions of the human hand, obtain the gender and age range of the personnel according to the human face recognition result;

[0056] Input the personnel position area into the human body recognition model, recognize the human body parts, take multiple key points for each part, and take the average value. Specifically, after the face temperature is recognized by the face, the nose tip point, left cheek, right cheek, and forehead are taken, and finally the average value is taken, and the calculation formula is:

[0057]

[0058] In the formula, T m is the face temperature, ℃; T mb is the temperature of the nose tip point, ℃; T mz is the left cheek temperature, ℃; T my is the right cheek temperature, ℃; T me is the forehead temperature, ℃.

[0059] After the hand temperature is recognized, the left hand back, right hand back, left wrist, and right wrist temperature are selected to take the average value, and the calculation formula is:

[0060]

[0061] In the formula, T s is the hand temperature, ℃; T sbz is the left hand back temperature, ℃; T sby is the right hand back temperature, ℃; T swz is the left wrist temperature, ℃; T swy is the right wrist temperature, ℃.

[0062] S24, continuously track the recognized human body parts, recognize the human body action, obtain the unit time moving distance, identify the human body activity state according to the human body action and the unit time moving distance of the human body parts, and estimate the indoor personnel load; wherein the indoor personnel load calculation formula is:

[0063] Wherein, Q p is the personnel load; Q z is the estimated thermal load of the zth population; n z is the number of the zth population; g is the total number of the population; the population includes male teenagers, male young people, male middle-aged people, male old people, female teenagers, female young people, female middle-aged people, and female old people;

[0064] More specifically, the personnel load is the heat load generated by each person in a specific environment, which is affected by various factors, so the present application only makes a simple estimate here to assist in rapid adjustment of the indoor environment, and the personnel load standard is shown in Table 1. Figure 3

[0065] S25, the pre-divided indoor floor surface temperature, indoor wall temperature, indoor surface temperature other than the floor surface and the wall are obtained from the thermal imaging video, and the indoor average radiation temperature and the operation temperature first calculation value are respectively calculated according to the indoor floor surface temperature, the indoor wall temperature and the indoor surface temperature other than the floor surface and the wall.

[0066] The calculation formula of the indoor average radiation temperature is as follows:

[0067]

[0068] Wherein, F j is the jth surface angle coefficient of the human body; T j is the absolute temperature of the jth surface; T r is the indoor average radiation temperature; k is the total number of surfaces; preferably, the human body pixels can be removed after the human body recognition algorithm is completed, and then the average radiation temperature of the remaining part is calculated,

[0069] The calculation formula of the operation temperature first calculation value (thermal comfort operation temperature) is as follows:

[0070]

[0071] Wherein, h c is the convective heat transfer coefficient; T a is the indoor air temperature; h r is the radiation heat transfer coefficient; T r is the average radiation temperature; T OP is the operation temperature first calculation value.

[0072] Specifically, when the thermal imager is installed, the image information is imported, the video information is corrected to eliminate lens distortion, and the noise is removed by a filter. In the graphical user interface (GUI), the radiant floor area, the wall, and the other indoor area are pre-divided.

[0073] Operation temperature: because the radiant floor in the radiant floor air conditioning system exchanges heat with the human body through radiation, the measurement standard of the indoor environment cannot be measured only by the indoor temperature and humidity, and the operation temperature is introduced. The operation temperature is the temperature of the comprehensive action of the indoor air temperature and the average radiation temperature on the human body.

[0074] As Figure 4 ​As shown, in step S3, a thermal sensation prediction model is established based on indoor environmental parameters and thermal comfort calculation data to predict the thermal sensation of indoor occupants. Specifically, this includes:

[0075] S31, using historical indoor humidity data, historical indoor temperature data, historical facial temperature data, historical hand temperature data, and historical indoor average radiant temperature data (which may also include historical data of the first calculated value of operating temperature) as the first input data, and using historical thermal sensation data of indoor personnel as the first output data, establish a first mapping database between the first input data and the corresponding first output data.

[0076] Specifically, a temperature and humidity recorder can be used in the laboratory to continuously measure the indoor thermal environment, and a thermal imager can be used to record the data, repeating step S2 above. The initial calculated values ​​of indoor temperature and humidity, facial temperature, hand temperature, average radiant temperature, and operating temperature are obtained. During the experiment, the indoor temperature and humidity, floor surface temperature, and wall temperature are adjusted at fixed intervals.

[0077] like Figure 5 As shown, 10 minutes after the environment was adjusted to the target environment, the subjects' thermal sensations were collected through a questionnaire, including: very cold, cold, cool, slightly cool, neutral, slightly warm, warm, hot, and very hot; finally, a first mapping database of the first input data and the corresponding first output data was established.

[0078] S32, using artificial neural network algorithms and the first mapping database, establish and train a thermal sensation prediction model;

[0079] Select input item: Facial temperature (T) m Hand temperature T s Indoor temperature T a Indoor humidity (RH) in First calculated value of operating temperature T op Indoor average radiant temperature T r The output item is thermal sensation.

[0080] Write an artificial neural network algorithm, divide the training set and test set, and normalize the input and output data;

[0081] Determine the neural network structure: the input layer has 6 neurons; try different numbers of neurons in the hidden layer; the output layer has 1 neuron. Input the hyperparameters of the model and start training. Complete the training of the thermal sensation prediction model.

[0082] The test set is input into the thermal sensation prediction model to predict thermal sensation. The model is evaluated using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). If the model is trained on the same dataset, the smaller the RMSE and MAE values, and the closer the R² is to 1, the higher the model accuracy.

[0083] It should be noted that the aforementioned artificial neural network can be based on algorithms such as recurrent neural networks and long short-term memory networks for prediction, and the embodiments of this application do not limit this.

[0084] S33 inputs real-time indoor temperature data, real-time indoor humidity data, real-time facial temperature data, real-time hand temperature data, and real-time indoor average radiant temperature data (which may also include real-time data of the first calculated value of operating temperature) into the trained thermal sensation prediction model to predict human thermal sensation.

[0085] Specifically, cold, cool, slightly cool, neutral, slightly warm, hot, and very hot are used as temperature regulation indicators, representing temperatures 2°C below the comfort temperature, 1°C below the comfort temperature, 0.5°C below the comfort temperature, equal to the comfort temperature, 0.5°C above the comfort temperature, 1°C above the comfort temperature, and 2°C below the comfort temperature, respectively.

[0086] Furthermore, the present invention does not limit the classification of temperature indicators according to actual conditions.

[0087] like Figure 6 As shown, in step S4, based on indoor and outdoor environmental parameters, equipment parameters, and air conditioning system operating parameters, an indoor environment and energy consumption prediction model is established to predict the indoor environment after the air conditioning system is adjusted and the energy consumption of the air conditioning system. Specifically, this includes:

[0088] S41, using historical data of wall temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air velocity, floor surface temperature, radiant floor water supply temperature, and indoor occupant load as the second input data, and using historical data of indoor average radiant temperature after air conditioning system adjustment, historical data of indoor condensation temperature difference after air conditioning system adjustment, historical data of air system energy consumption, and historical data of water system energy consumption as the second output data, establish a second mapping database between the second input data and the corresponding second output data;

[0089] Specifically, the historical data acquired by S1 is compiled. Historical air conditioning system operation data includes: fresh air temperature, fresh air velocity, radiant floor surface temperature, radiant floor supply water temperature, and radiant floor return water temperature; historical indoor environmental parameters include: indoor temperature, indoor humidity, wall temperature, and surface temperatures other than walls and floor surfaces; historical personnel activities include: number of personnel, personnel activity status, and personnel load; historical outdoor environmental parameters include: solar radiation, outdoor temperature, and outdoor humidity. In other words, historical data includes wall temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air velocity, floor surface temperature, radiant floor supply water temperature, and indoor personnel load.

[0090] The second output data includes historical data on the indoor average radiant temperature after air conditioning system adjustment, historical data on the indoor condensation temperature difference after air conditioning system adjustment, historical data on the energy consumption of the air system, and historical data on the energy consumption of the water system. The formula for calculating the condensation temperature difference is as follows:

[0091] ΔT l =T d -T ld

[0092] In the formula, T d T represents the floor surface temperature, in °C. ld ΔT represents the dew point temperature of the indoor air, in °C. l The temperature difference at which condensation occurs is expressed in °C.

[0093] The formula for calculating the energy consumption of a wind system is:

[0094] ΔE x =V×ρ s ×(C pa ×T o +w o (C pw ×T o +R w ×T o )-C pa ×T f +w f (C pw ×T f +R w ×T f )),

[0095] In the formula, V represents the air volume of the air system, in meters. 3 / h;ρ s C is the density of water, kg / m3; pa Specific heat capacity of dry air, kJ / kg·K; T o Outdoor air temperature, °C; w ofor outdoor air moisture content, kgH2O / kgDryAir; C pw for specific heat capacity of water vapor, kJ / kg·K; R W for gas constant of water vapor, kJ / kg·K; T f for fresh air fan outlet air temperature, ℃; w f for fresh air fan outlet air moisture content, kgH2O / kgDryAir; ΔE x for energy consumption of air system per hour, kW / h.

[0096] The water system energy consumption calculation formula is:

[0097] ΔE s =Q r ×ρ s ×C×ΔT

[0098] In the formula, ΔE s is the energy dissipated by the water system per hour, kW / h; Q r is the water flow, m 3 / h; ρ s is the water density, kg / m3; C is the specific heat capacity of water, about 4.18 kJ / (kg·K); ΔT r is the radiant floor supply and return water temperature difference, ℃.

[0099] S42, using artificial neural network algorithm and second mapping database, establishing and training indoor environment and energy consumption prediction model;

[0100] Select input items: wall temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air speed, floor surface temperature, radiant floor water supply temperature, personnel load; output items: indoor average radiant temperature historical data after air conditioning system adjustment, indoor dew point temperature difference historical data after air conditioning system adjustment, air system energy consumption historical data, water system energy consumption historical data.

[0101] Write artificial neural network algorithm; divide training set and test set, and normalize input and output data;

[0102] Determine the neural network structure, the number of input layer neurons is 8, try and fit the number of hidden layer neuron nodes, the number of output layer neurons is 4;

[0103] Input the hyperparameters of the model and start training; complete the training of indoor environment and energy consumption prediction model.

[0104] The test set is input into the indoor environment and energy consumption prediction model to predict the indoor environment and energy consumption, and the model is evaluated using the root mean square error (RMSE), mean absolute error (MAE), and determination coefficient R2. If the model is trained based on the same data set, the smaller the RMSE and MAE values obtained, and the closer R2 is to 1, the higher the accuracy of the model.

[0105] It should be noted that the artificial neural network algorithm described above can be based on a recurrent neural network, a long short-term memory network, or the like, and the embodiments of the present application do not limit this.

[0106] S43, input the wall surface temperature real-time data, outdoor temperature real-time data, outdoor humidity real-time data, fresh air temperature real-time data, fresh air speed real-time data, floor surface temperature real-time data, radiant floor supply water temperature real-time data, and indoor personnel load real-time data into the trained indoor environment and energy consumption prediction model to predict the indoor average radiant temperature real-time data after adjustment of the air conditioning system, the indoor dew point difference real-time data after adjustment of the air conditioning system, the air system energy consumption real-time data, and the water system energy consumption real-time data.

[0107] In step S5, the constraint conditions include an air supply amount constraint, a water supply pump state constraint, and a floor supply water temperature constraint; the air supply amount constraint condition is:

[0108]

[0109] wherein, is the minimum value of the fresh air system air supply amount; is the maximum value of the fresh air system air supply amount; is the regulated fresh air system air supply amount;

[0110] The water supply pump state constraint condition is:

[0111] 0≤K b ≤1,

[0112] wherein, 0 is off, 1 is on, K b is the water pump state;

[0113] The floor supply water temperature constraint condition is:

[0114]

[0115] wherein, is the minimum value of the floor supply water temperature; is the maximum value of the floor supply water temperature; is the regulated floor supply water pump temperature;

[0116] For example, Figure 7The personnel thermal comfort target function, the energy consumption target function are respectively constructed by indoor and outdoor environment parameters, indoor personnel load, thermal sensation and air conditioning system energy consumption, and the total target function is determined according to the weight of the personnel thermal comfort target function and the energy consumption target function, and the total target function specifically includes:

[0117] S51, the outdoor temperature, the outdoor humidity, the wall surface temperature, the floor surface temperature, the indoor temperature, the indoor humidity, the operation temperature, and the personnel load are taken as environmental independent variables, and the fresh air volume of the fresh air machine, the radiant floor water supply temperature, and the radiant floor water supply pump state are taken as control parameter independent variables.

[0118] Specifically, the real-time data set described in S1 is exported, including: wall surface temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air speed, floor surface temperature, radiant floor water supply temperature, and the personnel load obtained in S2 and the first calculation value of the operation temperature.

[0119] The outdoor temperature, the outdoor humidity, the wall surface temperature, the floor surface temperature, the indoor temperature, the indoor humidity, the first calculation value of the operation temperature, and the personnel load are taken as environmental data sets, and are taken as part of the independent variables, in turn X1, X2, X3, X4, X5, X6, X7, X8.

[0120] The control parameters are taken as part of the independent variables, that is, the fresh air volume of the fresh air machine, the radiant floor water supply temperature, and the radiant floor water supply pump state, in turn X9, X 10 , X 11 . That is, the control parameter independent variable is combined with the above-mentioned environmental parameter independent variable, and is taken as the independent variable of the optimization algorithm.

[0121] S52, the personnel thermal comfort, the personnel load, the condensation risk, the wind system energy consumption and the water system energy consumption are respectively taken as optimization targets.

[0122] f1=min|T OPS (X i )-T OP |,

[0123] f 1' =min|T in (X i )-T sd |,

[0124] f2=-min|ΔT l (X i )|,

[0125] f3=minE f (X i ),

[0126] f4=minE s (X i ),

[0127] f5 = min |Q p - |E f - E fs ||,

[0128] wherein, f1 is a thermal comfort target function; f 1' is a user set temperature target function; f2 is a dew point temperature difference target function; f3 is a wind system target function; f4 is a water system target function; f5 is a personnel load target function; X i is the ith independent variable; T OPS is an indoor operating temperature second calculation value calculated by an indoor average radiant temperature predicted by an indoor environment and energy consumption prediction model, ℃; T OP is an operating temperature first calculation value, ℃; T in is an indoor temperature predicted by an indoor environment and energy consumption prediction model, ℃; ΔT l is a dew point temperature difference predicted by an indoor environment and energy consumption prediction model, ℃; T sd is a user set temperature, ℃; E f is a wind system energy consumption predicted by an indoor environment and energy consumption prediction model, kw·h; E s is a water system energy consumption predicted by an indoor environment and energy consumption prediction model, kw·h; Q p is an indoor personnel load, w; E fs is a wind system energy consumption of a previous time period, kw·h;

[0129] With outdoor temperature, outdoor humidity, cooling load, humidification load, fresh air volume of fresh air fan, radiant floor water supply temperature and radiant floor water supply pump state as variables, set indoor air temperature, and based on a difference between the indoor air temperature and a user set value, construct an indoor temperature target function (i.e. user set temperature target function) f 1' ;

[0130] With outdoor temperature, outdoor humidity, wall surface temperature, floor surface temperature, indoor temperature, indoor humidity, operating temperature, personnel load, fresh air volume of fresh air fan, radiant floor water supply temperature and radiant floor water supply pump state as variables, construct a personnel thermal comfort target function f1;

[0131] With outdoor temperature, outdoor humidity, wall surface temperature, floor surface temperature, indoor temperature, indoor humidity, operating temperature, personnel load, fresh air volume of fresh air fan, radiant floor water supply temperature and radiant floor water supply pump state as variables, construct a dew point temperature difference target function f2;

[0132] The outdoor temperature, outdoor humidity, wall surface temperature, floor surface temperature, indoor temperature, indoor humidity, operation temperature, personnel load, fresh air volume of fresh air machine, radiation floor water supply temperature, and radiation floor water supply pump state are used as variables to construct a wind system energy consumption objective function f3.

[0133] The outdoor temperature, outdoor humidity, wall surface temperature, floor surface temperature, indoor temperature, indoor humidity, operation temperature, personnel load, fresh air volume of fresh air machine, radiation floor water supply temperature, and radiation floor water supply pump state are used as variables to construct a water system energy consumption objective function f4.

[0134] The indoor personnel load, wind system energy consumption predicted by the indoor environment and energy consumption prediction model, and wind system energy consumption of the previous period are used as variables to construct a water system energy consumption objective function f5.

[0135] Specifically, the objective function of the optimization algorithm is set, and the personnel thermal comfort, personnel load, dew risk, and energy consumption are used as optimization targets. The energy consumption is divided into wind system energy consumption and water system energy consumption. It is worth noting that when the number of indoor personnel changes, the thermal balance of the indoor environment is broken, and the air conditioning system control parameters need to be quickly changed to cope with the change in personnel load. Due to the thermal inertia of the radiant floor, it is difficult to handle instantaneous load changes, so short-term load changes are handed over to the wind system for processing, and the system is disabled after a period of time.

[0136] S53, according to the weight of each objective function, determine the total objective function f:

[0137] f = a1f1 + a2f2 + a3f3 + a4f4 + a5f5 + a6f 1' ,

[0138] Wherein, a1, a2, a3, a4, a5, a6 are the weights of each target.

[0139] Specifically, to reduce the complexity of the algorithm and simplify the multi-objective optimization problem, the indoor thermal and humid environment, dew risk, and energy consumption functions are weighted to form thermal, humid, dew, and energy consumption objective functions f. The weight of each objective function can be adjusted according to the actual needs of the user. Among them, a5 is not 0 only when the indoor personnel flow.

[0140] Wherein, as shown in Figure 8 The improved scarab optimization algorithm is applied to solve the total objective function to obtain the optimal control parameters of the floor radiant air conditioning system, which specifically includes:

[0141] The environmental independent variables and control parameter independent variables corresponding to the total objective function are used as a scarab, and the scarab population is initialized using Bernoulli mapping;

[0142] Specifically, the dung beetle optimization algorithm includes five behavior parts: rolling, dancing, breeding, foraging and stealing. The dung beetle population is initialized using Bernoulli mapping to increase population diversity and enhance the speed of the algorithm in traversing the global environment, thereby improving the optimization efficiency and convergence speed. The calculation formula is:

[0143]

[0144] wherein, Z n is the position of the nth dung beetle (the current value of the nth chaotic sequence generated), Z n+1 is the position of the n+1th dung beetle (the current value of the n+1th chaotic sequence generated). ρ is a mapping parameter, and its chaotic orbit state value range is (0, 1).

[0145] The fitness value of the dung beetle population is calculated, and the fitness value is the total objective function;

[0146] The current dung beetle position is updated, and the current fitness value is calculated for each updated dung beetle;

[0147] Specifically, updating the current dung beetle position specifically includes:

[0148] When no obstacle is encountered in the rolling process, when λ<γ, it is an obstacle-free state, otherwise it is an obstacle state, wherein λ is a random number, λ∈[0,1]; γ is a probability value, γ=0.9. At this time, the dung beetle position is updated as:

[0149]

[0150] wherein, t is the current iteration number; represents the position of the ith dung beetle at the tth iteration; α is a natural coefficient, taking 1, -1; k is a deflection coefficient constant, taking a value range (0, 0.2]; b is a constant between (0, 1); is the position of the worst dung beetle in the population;

[0151] When an obstacle is encountered in the rolling process, the dung beetle will re-determine the rolling direction θ by dancing. The rolling direction θ is between (0, π). After the rolling direction is re-determined, the dung beetle will continue to roll. At this time, the dung beetle position update formula is as follows:

[0152]

[0153] Since tanθ is 0 or does not exist when θ=0, π / 2, π, the dung beetle position is not updated at this time.

[0154] The dung beetle will roll the dung ball to a safe and suitable area for laying eggs and hide it. The boundary selection strategy is used to simulate the egg-laying area:

[0155]

[0156] wherein: Lb* and Ub* represent the lower and upper limits of the oviposition area, respectively, is the current local optimum position, R = 1 - t / t max is the linear convergence factor, t max is the maximum number of iterations, Lb and Ub are the lower and upper limits of the problem to be optimized, respectively;

[0157] When the oviposition area is determined, the breeding scarab beetle starts to lay eggs, and only one egg is laid at a time, and the position of the egg is also dynamically updated. At this time, the position update formula of the scarab beetle egg is:

[0158]

[0159] wherein, is the position of the s-th egg at the t-th iteration, b1 and b2 are both independent random vectors of 1 x D, and D represents the control parameter independent variable; that is, the dimension of the optimization problem, which is specifically: the fresh air volume of the new fan, the start-stop state of the radiant floor water supply pump, and the radiant floor water supply temperature.

[0160] When the larvae break the egg and grow into small scarab beetles, they will start foraging. At this time, the position update formula of the small scarab beetle foraging is:

[0161]

[0162] wherein, Lb l and Ub l are the lower and upper limits of the foraging range of the small scarab beetle, is the current global optimum position;

[0163] An adaptive T-distribution disturbance is introduced in the foraging stage of the small scarab beetle. t(iter) is a T-distribution variation disturbance with the iteration number iter as the degree of freedom parameter, which is used to disturb the foraging behavior of the small scarab beetle. At this time, the position update formula of the small scarab beetle is as follows:

[0164]

[0165] wherein, t(iter) is a T-distribution variation disturbance with the iteration number iter as the degree of freedom parameter;

[0166] There is a behavior of stealing dung balls in the scarab beetle population. At this time, the position update formula of the stealing scarab beetle is:

[0167]

[0168] wherein, S is a constant, and g is a random vector obeying a normal distribution.

[0169] In the population, the proportion of the rolling ball, the breeding scarab, the foraging scarab and the thief scarab can be 6:6:8:10, or can be adjusted according to actual conditions. The updated each scarab is judged, and the current fitness value and the optimal value are updated and calculated.

[0170] It is determined whether a preset stop condition is met, and the preset stop condition includes reaching a preset maximum iteration number.

[0171] As shown in Figure 9 The intelligent control system of the floor radiation air conditioning system provided by the technical scheme of the present application further comprises:

[0172] S7, based on the optimal control parameter of the floor radiation air conditioning system control and the indoor environment parameter, the floor radiation air conditioning system is controlled.

[0173] As shown in Figure 10 S7 specifically includes:

[0174] It is determined whether the difference between the floor surface temperature and the dew point temperature is greater than a preset difference threshold (for example, 1℃), and if not, the fresh air volume of the fresh air fan is adjusted to the highest gear;

[0175] If it is greater, it is determined whether the fresh air volume of the fresh air fan is not less than the minimum fresh air volume required by the indoor personnel, and if not, the floor radiation air conditioning system is controlled based on the optimal control parameter of the floor radiation air conditioning system control; if it is less, the constraint condition is changed, and the improved scarab optimization algorithm is applied again to solve the optimal control parameter, until the fresh air volume of the fresh air fan is not less than the minimum fresh air volume required by the indoor personnel.

[0176] When solving the optimal control parameter, the anti-condensation and minimum fresh air volume requirements are required. Specifically, the control logic of the air conditioning system anti-condensation and pre-dehumidification, before starting the anti-condensation radiation cooling equipment, when the condensation temperature difference in the room is less than 1℃, the radiation floor surface has a condensation risk, the fresh air volume of the fresh air fan is adjusted to the highest gear, effectively reducing the humidity of the indoor air, reducing the condensation risk, and avoiding the condensation problem.

[0177] Furthermore, the minimum fresh air volume, in the step S2, the formula for calculating the minimum fresh air volume Q required by the indoor personnel according to the number of indoor activities N is as follows: Q=q×N;

[0178] Q=q×N

[0179] Wherein, q is the minimum fresh air volume required by each person; N is the number of indoor activities; Q is the minimum fresh air volume required by the indoor personnel;

[0180] Specifically, for an office, q is 30m 3 / (h·person), and for a conference room, q is 11-14m3 / (h·person), for a classroom, q is 22-28 m 3 / (h·person).

[0181] When the air conditioning system is running, first, indoor and outdoor environment parameters, thermal imaging image data, and ordinary image data are acquired; computer vision technology is used to process the images to acquire the number of personnel, personnel load, face temperature, hand temperature, average radiant temperature, operating temperature, and floor surface temperature; then, the risk of condensation is judged, i.e., whether the floor surface temperature is higher than the dew point temperature by 1 DEG C; if not, the fresh air volume of the fresh air fan is adjusted to the highest gear to rapidly dehumidify the indoor air, and the current air conditioning system control process is completed; if yes, the improved scarab beetle optimization algorithm is used to solve the optimal control parameters; further, it is judged whether the fresh air volume of the fresh air fan is greater than or equal to the minimum fresh air volume required by the indoor personnel; if not, the constraint condition is changed to meet the minimum fresh air volume required by the indoor personnel, the improved scarab beetle optimization algorithm is used to solve the optimal control parameters and control the air conditioning system according to the optimal control parameters, and the current air conditioning system control process is completed; if yes, the air conditioning system is controlled according to the optimal control parameters, and the current air conditioning system control process is completed.

[0182] In the technical scheme of the present application, computer vision technology is used to process thermal imaging video streams and non-thermal imaging video streams, track indoor personnel, acquire indoor personnel load estimation data and thermal comfort calculation data, estimate indoor personnel load according to the indoor personnel load estimation data, establish a thermal sensation prediction model based on indoor environment parameters and thermal comfort calculation data to predict indoor personnel thermal sensation, establish an indoor environment and energy consumption prediction model based on indoor and outdoor environment parameters, equipment parameters, and air conditioning system operation parameters to predict the indoor environment after air conditioning system adjustment and air conditioning system energy consumption, determine constraint conditions, construct personnel thermal comfort objective functions and energy consumption objective functions respectively based on indoor and outdoor environment parameters, indoor personnel load, thermal sensation, and air conditioning system energy consumption, determine a total objective function according to the weights of the personnel thermal comfort objective functions and the energy consumption objective functions, apply the improved scarab beetle optimization algorithm to solve the total objective function to obtain optimal control parameters for floor radiant air conditioning system control, which can adjust radiant air conditioning system parameters in real time according to indoor personnel load and personnel thermal comfort, effectively solve the problem of low reliability of floor radiant air conditioning system control caused by existing technologies, and effectively improve the reliability of floor radiant air conditioning system control.

[0183] The technical scheme of the present application establishes a personnel detection and tracking model based on computer vision technology, obtains the position and number of detected personnel, obtains the gender and age range of personnel according to the human face recognition result, continuously tracks the recognized human body parts, recognizes human body actions, obtains the moving distance per unit time, identifies the human body activity state according to the human body action and the moving distance per unit time of the human body parts, estimates the indoor personnel load, and estimates the indoor personnel load according to the personnel type and activity state, thereby further improving the reliability of the floor radiant air conditioning system control.

[0184] In the technical scheme of the present application, the indoor and outdoor environment parameters and thermal imaging data are monitored in real time, the artificial neural network algorithm is used, the thermal sensation prediction model is established based on the indoor environment parameters and thermal comfort calculation data, the indoor personnel thermal sensation is predicted, and the user thermal sensation is accurately reflected; personalized intelligent adjustment of the radiant air conditioning system can be realized, the user's cumbersome operation is reduced, and the user's thermal comfort is greatly improved.

[0185] In the technical scheme of the present application, the computer vision tracking is used to identify the indoor personnel number and activity, the indoor personnel load change is responded in time, the air conditioning parameters are quickly adjusted, the hysteresis effect in the control of the radiant air conditioning system is effectively alleviated, and the discomfort and energy waste caused by the imbalance between air conditioning load supply and demand are reduced.

[0186] In the technical scheme of the present application, the artificial neural network algorithm is used to predict the regulation and control effect of the radiant air conditioning system, and the constraint conditions are determined, the personnel thermal comfort target function and the energy consumption target function are respectively constructed by using the indoor and outdoor environment parameters, the indoor personnel load, the thermal sensation and the air conditioning system energy consumption, the total target function is determined according to the weight of the personnel thermal comfort target function and the energy consumption target function, the improved scarab beetle optimization algorithm is applied, and the optimal regulation and control parameters are generated, so that the personnel thermal comfort is ensured and the air conditioning system energy consumption is reduced.

[0187] Embodiment two

[0188] As shown in Figure 11 The technical scheme of the present application further provides a floor radiant air conditioning system intelligent control system, which comprises:

[0189] The first acquisition module 101 acquires indoor and outdoor environment parameters, thermal imaging video streams, non-thermal imaging video streams, air conditioning system operation data and equipment parameters.

[0190] The second acquisition module 102 processes the thermal imaging video streams and the non-thermal imaging video streams using the computer vision technology, tracks the indoor personnel, acquires indoor personnel load estimation data and thermal comfort calculation data, and estimates the indoor personnel load according to the indoor personnel load estimation data.

[0191] The first establishing module 103 establishes a thermal sensation prediction model based on the indoor environment parameters and the thermal comfort calculation data, and predicts indoor personnel thermal sensation;

[0192] The second establishing module 104 establishes an indoor environment and energy consumption prediction model based on the indoor and outdoor environment parameters, the equipment parameters and the air conditioning system operation parameters, and predicts the indoor environment after adjustment of the air conditioning system and the energy consumption of the air conditioning system;

[0193] The determining module 105 determines constraint conditions, constructs personnel thermal comfort target functions and energy consumption target functions respectively based on the indoor and outdoor environment parameters, indoor personnel load, thermal sensation and air conditioning system energy consumption, and determines a total target function according to the weights of the personnel thermal comfort target functions and the energy consumption target functions;

[0194] The solving module 106 applies an improved Canthon optimization algorithm to solve the total target function, and obtains optimal control parameters of the floor radiant air conditioning system.

[0195] It should be noted that the implementation processes in the first obtaining module 101, the second obtaining module 102, the first establishing module 103, the second establishing module 104, the determining module 105 and the solving module 106 correspond to the method steps in Embodiment 1, and thus will not be described herein.

[0196] In the technical scheme, the computer vision technology is used to process thermal imaging video streams and non-thermal imaging video streams, track indoor personnel, obtain indoor personnel load estimation data and thermal comfort calculation data, estimate indoor personnel load according to the indoor personnel load estimation data, establish a thermal sensation prediction model based on the indoor environment parameters and the thermal comfort calculation data, predict indoor personnel thermal sensation, establish an indoor environment and energy consumption prediction model based on the indoor and outdoor environment parameters, the equipment parameters and the air conditioning system operation parameters, predict the indoor environment after adjustment of the air conditioning system and the energy consumption of the air conditioning system, determine constraint conditions, construct personnel thermal comfort target functions and energy consumption target functions respectively based on the indoor and outdoor environment parameters, indoor personnel load, thermal sensation and air conditioning system energy consumption, determine a total target function according to the weights of the personnel thermal comfort target functions and the energy consumption target functions, and apply an improved Canthon optimization algorithm to solve the total target function, so as to obtain optimal control parameters of the floor radiant air conditioning system, adjust radiant air conditioning system parameters in real time according to indoor personnel load and personnel thermal comfort, effectively solve the problem of low reliability of floor radiant air conditioning system control caused by the prior art, and effectively improve the reliability of floor radiant air conditioning system control.

[0197] In the technical scheme of the present application, a personnel detection and tracking model is established based on computer vision technology to obtain the position and number of detected personnel, the gender and age range of personnel are obtained according to the human face recognition result, the human body parts are continuously tracked, the human body actions are recognized, the moving distance per unit time is obtained, the human body activity state is recognized according to the human body actions and the moving distance per unit time of the human body parts, the indoor personnel load is estimated, the indoor personnel compliance can be estimated according to the personnel type and the activity state, and the reliability of the floor radiant air conditioning system control is further improved.

[0198] In the technical scheme of the present application, the indoor and outdoor environmental parameters and thermal imaging data are monitored in real time, a thermal sensation prediction model is established based on the indoor environmental parameters and thermal comfort calculation data by using an artificial neural network algorithm, the indoor personnel thermal sensation is predicted, and the user thermal sensation is accurately reflected; personalized intelligent adjustment of the radiant air conditioning system can be realized, the user's cumbersome operation is reduced, and the user thermal comfort is greatly improved.

[0199] In the technical scheme of the present application, the computer vision tracking is used to identify the indoor personnel number and activity, the indoor personnel load change is responded in time, the air conditioning parameters are quickly adjusted, the hysteresis effect in the control of the radiant air conditioning system is effectively alleviated, and the discomfort and energy waste caused by the imbalance between the air conditioning load supply and demand are reduced.

[0200] In the technical scheme of the present application, the artificial neural network algorithm is used to predict the regulation and control effect of the radiant air conditioning system, and the constraint conditions are determined, the personnel thermal comfort target function and the energy consumption target function are respectively constructed by using the indoor and outdoor environmental parameters, the indoor personnel load, the thermal sensation and the air conditioning system energy consumption, the total target function is determined according to the weight of the personnel thermal comfort target function and the energy consumption target function, the improved scarab beetle optimization algorithm is applied to generate the optimal regulation and control parameters, so as to ensure the personnel thermal comfort and reduce the energy consumption of the air conditioning system.

[0201] Although the specific embodiments of the present application are described above with reference to the accompanying drawings, the present application is not limited to the above description, and various modifications or changes can be made by those skilled in the art without creative labor on the basis of the technical scheme of the present application.

Claims

1. A method for intelligent control of a floor radiant air conditioning system, characterized in that, include: Acquire indoor environmental parameters, outdoor environmental parameters, thermal imaging video streams, non-thermal imaging video streams, and air conditioning system operation data. The air conditioning system operation data includes: fresh air temperature, fresh air velocity, fresh air unit airflow, fresh air unit fresh air volume, radiant floor surface temperature, radiant floor supply water temperature, and radiant floor return water temperature. Indoor environmental parameters include: indoor temperature, indoor humidity, wall temperature, and surface temperatures other than walls and floor surfaces. Outdoor environmental parameters include: solar radiation, outdoor temperature, and outdoor humidity. Indoor occupant load estimation data includes the number of people, their gender, age range, and activity status. Thermal comfort calculation data includes facial temperature, hand temperature, average indoor radiant temperature, and operating temperature, where the operating temperature is the combined effect of indoor air temperature and average indoor radiant temperature on the human body. Computer vision technology is used to process thermal imaging video streams and non-thermal imaging video streams, track indoor occupants, obtain indoor occupant load estimation data and thermal comfort calculation data, and estimate indoor occupant load based on the indoor occupant load estimation data. Based on indoor environmental parameters and thermal comfort calculation data, a thermal sensation prediction model is established to predict the thermal sensation of people indoors. Based on indoor and outdoor environmental parameters, equipment parameters, and air conditioning system operating parameters, an indoor environment and energy consumption prediction model is established to predict the indoor environment and air conditioning system energy consumption after the air conditioning system is adjusted. Define the constraints, construct the objective function for thermal comfort and the objective function for energy consumption based on indoor and outdoor environmental parameters, indoor occupant load, thermal sensation, and air conditioning system energy consumption, respectively, and determine the overall objective function based on the weights of the objective functions for thermal comfort and energy consumption. An improved dung beetle optimization algorithm is applied to solve the overall objective function, thereby obtaining the optimal control parameters for the floor radiant air conditioning system.

2. The intelligent control method for a floor radiant air conditioning system according to claim 1, characterized in that, Computer vision technology is used to process thermal imaging video streams and non-thermal imaging video streams, track indoor occupants, and obtain indoor occupant load calculation data and thermal comfort calculation data. Based on the indoor occupant load estimation data, the specific indoor occupant load estimation includes: Based on computer vision technology, establish personnel detection model, personnel tracking model, and human body recognition model. Indoor thermal imaging video stream and non-thermal imaging video stream are both input into the trained personnel detection model, personnel tracking model, and human body recognition model. The indoor thermal imaging video stream and non-thermal imaging video stream are normalized, and the trained personnel detection model and personnel tracking model are used to identify and obtain the location and number of detected personnel. The location of the person is input into the trained human body recognition model, which identifies the body parts. The facial temperature is obtained based on the temperature values ​​of different parts of the face; the hand temperature is obtained based on the temperature values ​​of different parts of the hands; and the gender and age range of the person are obtained based on the facial recognition results. Continuous tracking of identified human body parts, recognition of human movements, acquisition of distance traveled per unit time, and identification of human activity status based on human movements and distance traveled per unit time of body parts to estimate indoor occupant load; the formula for calculating indoor occupant load is as follows: ,in, For personnel load; The estimated heat load for the z-th population group; Let g be the number of the z-th group; g is the total number of people; the group includes male teenagers, male youths, male middle-aged men, male elderly men, female teenagers, female youths, female middle-aged women, and female elderly women; The indoor floor surface temperature, indoor wall temperature, and other indoor surface temperatures are obtained from the thermal imaging video. Based on the indoor floor surface temperature, indoor wall temperature, and other indoor surface temperatures besides the floor and walls, the indoor average radiant temperature and the first calculated value of the operating temperature are calculated respectively.

3. The intelligent control method for a floor radiant air conditioning system according to claim 2, characterized in that, The formula for calculating the average indoor radiant temperature is: ， in, Let be the angle coefficient of the j-th surface relative to the human body; T represents the absolute temperature of the j-th surface. r k is the total number of surfaces; The formula for calculating the first calculated value of operating temperature is: ， in, The convective heat transfer coefficient; Indoor air temperature; The radiative heat transfer coefficient; The mean radiation temperature; This is the first calculated value for the operating temperature.

4. The intelligent control method for a floor radiant air conditioning system according to claim 2, characterized in that, Based on indoor environmental parameters and thermal comfort calculation data, a thermal sensation prediction model is established to predict the specific thermal sensations of indoor occupants, including: Using historical data on indoor humidity, indoor temperature, facial temperature, hand temperature, and indoor average radiant temperature as the first input data, and historical data on the thermal sensation of people indoors as the first output data, a first mapping database between the first input data and the corresponding first output data is established. A thermal sensation prediction model was established and trained using artificial neural network algorithms and a first mapping database. By inputting real-time indoor temperature data, real-time indoor humidity data, real-time facial temperature data, real-time hand temperature data, and real-time indoor average radiant temperature data into a trained thermal sensation prediction model, the human body's thermal sensation can be predicted.

5. The intelligent control method for a floor radiant air conditioning system according to claim 2, characterized in that, Based on indoor and outdoor environmental parameters, equipment parameters, and air conditioning system operating parameters, an indoor environment and energy consumption prediction model is established to predict the indoor environment after air conditioning system adjustment and the specific energy consumption of the air conditioning system, including: Using historical data on wall temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air velocity, floor surface temperature, radiant floor water supply temperature, and indoor occupant load as the second input data, and using historical data on indoor average radiant temperature after air conditioning system adjustment, historical data on indoor condensation temperature difference after air conditioning system adjustment, historical data on air system energy consumption, and historical data on water system energy consumption as the second output data, a second mapping database between the second input data and the corresponding second output data is established. An indoor environment and energy consumption prediction model was established and trained using artificial neural network algorithms and a second mapping database. Real-time data on wall temperature, outdoor temperature, outdoor humidity, fresh air temperature, fresh air velocity, floor surface temperature, radiant floor water supply temperature, and indoor occupant load are input into a trained indoor environment and energy consumption prediction model to predict real-time data on indoor average radiant temperature after air conditioning system adjustment, real-time data on indoor condensation temperature difference after air conditioning system adjustment, real-time data on air system energy consumption, and real-time data on water system energy consumption.

6. The intelligent control method for a floor radiant air conditioning system according to claim 5, characterized in that, The constraints include air supply volume constraints, water supply pump status constraints, and floor water supply temperature constraints; among them, the air supply volume constraint is: ， in, This represents the minimum air volume supplied by the fresh air system. This represents the maximum air volume supplied by the fresh air system. To regulate the air supply volume of the fresh air system; The water supply pump status constraints are as follows: ， Where 0 represents off and 1 represents on. Water pump status; The floor water supply temperature constraints are as follows: ， in, This is the lowest possible floor water temperature. This is the highest possible floor water temperature. To regulate the temperature of the floor water supply pump; Using indoor and outdoor environmental parameters, indoor occupant load, thermal sensation, and air conditioning system energy consumption, respectively construct objective functions for occupant thermal comfort and energy consumption. Based on the weights of the occupant thermal comfort and energy consumption objective functions, determine the specific components of the overall objective function: Outdoor temperature, outdoor humidity, wall temperature, floor surface temperature, indoor temperature, indoor humidity, operating temperature, and personnel load are used as environmental independent variables, while fresh air volume of fresh air unit, radiant floor water supply temperature, and radiant floor water supply pump status are used as control parameter independent variables. The optimization objectives are personnel thermal comfort, personnel load, condensation risk, air system energy consumption, and water system energy consumption, respectively. ， ， ， ， ， ， in, The objective function is thermal comfort. Set the temperature target function for the user; Let the condensation temperature difference be the objective function; Let the objective function be the wind system. Let the objective function of the water system be... The objective function for personnel load; X i ; The second calculated value of the indoor operating temperature is calculated based on the indoor average radiant temperature predicted by the indoor environment and energy consumption prediction model. First calculated operating temperature; T in The indoor temperature predicted by the indoor environment and energy consumption prediction model; The condensation temperature difference is predicted by an indoor environment and energy consumption prediction model; Set the temperature for the user; Energy consumption of the wind system predicted by the indoor environment and energy consumption prediction model; Water system energy consumption predicted by indoor environment and energy consumption prediction model; Q p For indoor occupancy load; E fs This represents the energy consumption of the wind system in the previous period. Based on the weights of each objective function, determine the overall objective function f: ， Among them, α1, α2, α3, α4, α5, The weights of each objective.

7. The intelligent control method for a floor radiant air conditioning system according to claim 1, characterized in that, By applying the improved dung beetle optimization algorithm, the overall objective function is solved, and the optimal control parameters for the floor radiant air conditioning system are obtained, including: Treat the environmental variables and regulation parameters corresponding to the overall objective function as a dung beetle, and initialize the dung beetle population using Bernoulli mapping; Calculate the fitness value of the dung beetle population, where the fitness value is the overall objective function; Update the current dung beetle position and calculate the current fitness value for each dung beetle after the update; Determine whether a preset stopping condition is met, the preset stopping condition including reaching a preset maximum number of iterations.

8. A method for intelligent control of a floor radiant air conditioning system according to any one of claims 1-7, characterized in that, it further includes... include: The floor radiant air conditioning system is regulated based on the optimal control parameters and indoor environmental parameters.

9. The intelligent control method for a floor radiant air conditioning system according to claim 8, characterized in that, Based on the optimal control parameters and indoor environmental parameters of the floor radiant air conditioning system, the specific control of the floor radiant air conditioning system includes: Determine if the difference between the floor surface temperature and the dew point temperature is greater than the preset difference threshold. If it is not greater, adjust the air volume of the fresh air unit to the highest level. If the value is greater than the minimum fresh air volume required by the indoor occupants, determine whether the fresh air volume of the fresh air unit is not less than the minimum fresh air volume required by the indoor occupants. If it is not less than the minimum fresh air volume required by the indoor occupants, adjust the floor radiant air conditioning system based on the optimal control parameters. If the value is less than the minimum fresh air volume required by the indoor occupants, change the constraints and reapply the improved dung beetle optimization algorithm to solve for the optimal control parameters until the fresh air volume of the fresh air unit is not less than the minimum fresh air volume required by the indoor occupants.

10. An intelligent control system for a floor radiant air conditioning system, characterized in that, include: The first acquisition module acquires indoor and outdoor environmental parameters, thermal imaging video streams, non-thermal imaging video streams, air conditioning system operation data, and equipment parameters. The air conditioning system operation data includes: fresh air temperature, fresh air velocity, fresh air unit airflow, fresh air unit fresh air volume, radiant floor surface temperature, radiant floor supply water temperature, and radiant floor return water temperature. Indoor environmental parameters include: indoor temperature, indoor humidity, wall temperature, and surface temperatures other than walls and floor surfaces. Outdoor environmental parameters include: solar radiation, outdoor temperature, and outdoor humidity. Indoor occupant load estimation data includes the number of people, their gender, age range, and activity status. Thermal comfort calculation data includes facial temperature, hand temperature, average indoor radiant temperature, and operating temperature, where the operating temperature is the combined effect of indoor air temperature and average indoor radiant temperature on the human body. The second acquisition module uses computer vision technology to process thermal imaging video streams and non-thermal imaging video streams, tracks indoor occupants, acquires indoor occupant load estimation data and thermal comfort calculation data, and estimates indoor occupant load based on the indoor occupant load estimation data. The first module establishes a thermal sensation prediction model based on indoor environmental parameters and thermal comfort calculation data to predict the thermal sensation of people indoors. The second module establishes an indoor environment and energy consumption prediction model based on indoor and outdoor environmental parameters, equipment parameters, and air conditioning system operating parameters, and predicts the indoor environment and air conditioning system energy consumption after the air conditioning system is adjusted. The modules and constraints are determined. The objective functions for human thermal comfort and energy consumption are constructed based on indoor and outdoor environmental parameters, indoor occupant load, thermal sensation, and air conditioning system energy consumption, respectively. The overall objective function is determined based on the weights of the human thermal comfort objective function and the energy consumption objective function. The solution module applies an improved dung beetle optimization algorithm to solve the overall objective function and obtain the optimal control parameters for the floor radiant air conditioning system.

Images