Airport terminal air conditioning system double-layer optimization control method considering passenger flow space-time performance

Abstract

The invention discloses an airport terminal air conditioning system double-layer optimization control method considering passenger flow space-time performance, and belongs to the field of air conditioner multi-area control. According to the method, a regional heat and humidity balance model and an air quality model which consider the passenger occupancy rate are constructed by dividing the functional regions of the terminal building. Based on a Markov chain and event passenger movement model, a transition probability matrix is utilized to describe the flow condition of passengers in each area, an occupied area and an unoccupied area are divided, and a differential control strategy is implemented for the areas with different occupied states. Compared with a traditional control strategy, a dynamic partition control mechanism is introduced, the problems of local pollutant accumulation and heat and humidity environment imbalance caused by uneven passenger flow distribution are effectively solved, the operation energy consumption of the air conditioning system is remarkably reduced, the balance of energy efficiency and comfort is achieved, and a new technical path is provided for optimization of the air conditioning system of the terminal building.

Description

Technical Field

[0001] The present invention belongs to the field of multi - zone control of air conditioners, and particularly relates to a double - layer optimal control method for an air - conditioning system in a terminal building considering the spatio - temporal characteristics of passenger flow. Background Art

[0002] As one of the main fields of energy consumption, the construction industry accounts for about one - third of the global total energy consumption. As a large - scale public transportation building, the annual average energy consumption per unit area of an airport terminal is about 180 kW·h / m 2 , which is about 2 - 3 times that of ordinary public buildings. Moreover, the energy consumption of the heating, ventilation, and air - conditioning (HVAC) system accounts for as high as 40% - 80%, becoming the key object of energy - saving optimization in the terminal building.

[0003] Terminal buildings have characteristics such as large - scale space, complex functional zoning, and dynamic changes in the regional environment, which pose higher requirements for the operation control of HVAC systems. Due to the large internal space span of the terminal building, the non - continuity of flight operations, and the spatio - temporal heterogeneity of passenger flow, the distribution of passengers in different regions is significantly uneven. The occupancy rate of different regions is affected by multiple factors such as flight schedules and passenger travel behaviors, and there are significant differences in the comfort requirements of different functional regions (such as check - in areas, security check areas, and waiting areas).

[0004] However, most of the existing air - conditioning control methods for terminal buildings adopt unified set parameters, failing to fully consider the dynamic changes in the occupancy rate of passengers in different regions, resulting in widespread energy waste. In addition, traditional control strategies are difficult to effectively cope with complex personnel flow and environmental changes, easily leading to problems such as local thermal - humidity environment imbalance and regional pollutant accumulation, further exacerbating the contradiction between energy consumption and environmental quality. As a multi - functional composite large - scale space building, the requirements for regional environmental control in the terminal building are significantly different from those of traditional public buildings, and its energy consumption characteristics and energy - saving potential have important research value. By constructing an air - conditioning control strategy based on zonal dynamic optimization, not only can the comfort requirements of different functional regions be accurately met, but also the energy consumption of the air - conditioning system can be significantly reduced, achieving the energy - saving and emission - reduction goals of the civil aviation industry. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects in the prior art and provide a double - layer optimal control method for an air - conditioning system in a terminal building considering the spatio - temporal characteristics of passenger flow. By quantitatively analyzing the dynamic distribution law of passengers in each functional region, the terminal building space is divided into occupied areas and unoccupied areas, and different control strategies are implemented for different regions. Through the double - layer optimization of the thermal - humidity control layer and the fresh - air control layer, on the premise of ensuring the thermal - humidity comfort and air quality in the terminal building, the operation energy consumption of the air - conditioning system can be effectively reduced.

[0006] The specific technical solutions adopted by the present invention are as follows:

[0007] The present invention provides a double - layer optimal control method for the terminal air - conditioning system considering the spatio - temporal characteristics of passenger flow. Based on the passenger movement mathematical model, the air - supply areas in each functional area of the terminal are divided into occupied areas and unoccupied areas, and different control strategies are implemented for different air - supply areas. The double - layer model predictive control algorithm is adopted, comprehensively considering the dynamic changes of passenger flow and the uncertainty of outdoor meteorological conditions, combined with the real - time state information of the system. Through the collaborative optimization of the heat - humidity control layer and the fresh - air control layer, the decision variables within the prediction horizon are solved iteratively, including the optimal control sequences of supply air temperature, supply air humidity, supply air volume flow rate, and fresh - air ratio, and the corresponding control instructions are executed.

[0008] Preferably, the control method is as follows:

[0009] S1: Define the division principle of the terminal air - conditioning control area, and construct a regional heat - humidity balance model and a regional air quality model considering passenger occupancy rate;

[0010] S2: Construct a passenger movement model based on Markov chain and events, describe the distribution characteristics of passengers in the area through the transition probability matrix, and divide the occupied areas and unoccupied areas of each area in real - time;

[0011] S3: Based on the regional heat - humidity balance model and regional air quality model in S1, construct the optimization objective function of the double - layer control model, and clarify the optimization objectives of the area under different passenger distribution characteristics;

[0012] S4: Set the parameters of the double - layer control model in S3, clarify the constraint conditions, and obtain the data within the prediction horizon;

[0013] S5: Based on the passenger movement model in S2 and the results of S3, construct a double - layer control model for the occupied area and the unoccupied area. The upper - layer heat - humidity control layer optimizes the supply air temperature, supply air humidity, and supply air volume flow rate according to the passenger density, outdoor weather conditions, and system state information, and the lower - layer fresh - air control layer optimizes the fresh - air ratio according to the regional air quality requirements;

[0014] S6: Execute the control instructions within the control horizon, update the state information of the double - layer control model in S5, and feedback the updated state information to the next control horizon. Repeat steps S2 - S5 and perform rolling optimization until the scheduling period ends.

[0015] Preferably, in S1, the division principle of the terminal air - conditioning control area includes dividing the terminal area into a check - in area, a security check area, and a waiting area according to the functional differences of different areas, and further dividing each of the three main functional areas into several sub - areas according to the air - supply areas to formulate air - conditioning control methods for different areas.

[0016] Preferably, in S1, the regional heat and moisture balance model includes a regional heat balance model and a regional moisture balance model, and the construction method is as follows:

[0017] For the regional heat balance model, it is calculated by the change value of the sensible heat of the regional air being equal to the difference between the regional sensible heat load and the air-conditioning cooling capacity, and is expressed as:

[0018]

[0019] Among them, the regional sensible heat load includes internal heat sources and external heat sources. The internal heat sources include passengers, equipment, and lighting, and the external heat sources include the building envelope, solar radiation, and infiltration; C p represents the specific heat capacity of the regional air, kJ / (kg·K); ρ a represents the density of the regional air, kg / m 3 ; V represents the volume of the regional space, m 3 ; T in represents the temperature of the regional air, °C; t represents the current time, s; T oa represents the outdoor air temperature, °C; j represents the type of the building envelope of the region; s represents the set of the building envelopes of the region; A j represents the area of the building envelope j, m 2 ; k j represents the heat transfer coefficient of the building envelope j, W / (m 2 ·°C); G sa represents the supply air volume flow rate, m 3 / s; T sa represents the supply air temperature, °C; Q person represents the sensible heat dissipation of passengers, kW; Q other represents the heat generation of other heat sources, kW;

[0020] Q other =Q solar +Q infilt +Q device +Q light (2)

[0021]

[0022] Among them, Q solar is the heat gain from solar radiation, kW; Q infilt is the heat transfer through air infiltration, kW; Q device is the heat dissipation of equipment, kW; Q light is the heat dissipation of lighting, kW; Q person represents the sensible heat dissipation of passengers, kW; N represents the number of passengers in the region, person;

[0023] For the regional moisture balance model, it is calculated by the change value of the latent heat of the regional air being equal to the difference between the regional latent heat load and the air-conditioning dehumidification amount, and is expressed as:

[0024]

[0025] M infilt =A inf Vρ a (W oa -W in ) (6)

[0026] Among them, the regional latent heat load includes infiltration and fresh air moisture load, and the latent heat dissipation load of passengers; W in represents the regional air humidity, g / kg; M person represents the latent heat dissipation load of passengers, g / s; M infilt represents the moisture load generated by air infiltration, g / s; W sa represents the supply air humidity, g / kg; N represents the number of passengers in the region, person; Δh vap represents the specific enthalpy of evaporation, 2257 J / g for water; A inf represents the infiltration air rate, s -1 ; W oa represents the outdoor air humidity, g / kg.

[0027] Preferably, in the S1, the construction method of the regional air quality model is specifically as follows:

[0028] The carbon dioxide concentration is used to measure the regional air quality. According to the mass balance equation, the regional carbon dioxide concentration balance model is obtained:

[0029]

[0030] C ma =drC oa +(1 - dr)C in (8)

[0031] G in =G person N (9)

[0032] Among them, C in represents the regional carbon dioxide volume fraction, ppm; C ma represents the mixed air carbon dioxide volume fraction, ppm; C oa represents the outdoor carbon dioxide volume fraction, ppm; G sa represents the regional supply air volume flow rate, m 3 / s; V represents the regional space volume, m 3 ; A inf represents the infiltration air rate, s -1; dr represents the fresh air ratio; G in represents the total carbon dioxide generation rate of passengers in the area, m 3 / s; G person represents the carbon dioxide generation rate of passengers, m 3 / (s·person); N represents the number of passengers in the area, person;

[0033] The carbon dioxide release rate per passenger G person The calculation formula is expressed as:

[0034]

[0035] Among them, A D represents the DuBois surface area, m 2 ; M is the physical activity level in Met, which is set according to the activity types of passengers in the terminal; RQ represents the respiratory quotient, which is the molecular ratio of CO2 released and O2 absorbed during respiration.

[0036] Preferably, in S2, passengers are regarded as individual mobile units, and by defining the transition probability matrix under different event mechanisms for their movement modes, the displacement of passengers in the area is characterized and simulated to generate the hourly occupancy of passengers in sub-areas, specifically as follows:

[0037] S21: Taking the average value of the predicted number of passengers on the day as the standard, define the event mechanism based on the passenger flow;

[0038] S22: Obtain the number of passengers in the three different functional areas of the check-in area, security check area, and waiting area under the current control time domain, and based on the event mechanism defined in S21, obtain the passenger transition probability matrix;

[0039] S23: Based on the passenger flow under the current control time domain and the passenger transition probability matrix in S22, calculate the passenger distribution in different air-conditioning supply areas in each functional area according to the Markov movement model, and divide the occupied area and the unoccupied area.

[0040] Preferably, S21 is specifically as follows:

[0041] If the number of passengers N(t) in the functional area under the current control time domain is less than the average number of passengers on the day then set the transfer probability p ij of half of the air-conditioning supply areas of the personnel to zero, and generate probability distribution random numbers for the remaining air-conditioning supply areas through the Monte Carlo method, and obtain the passenger number distribution in each sub-area according to the initial state and the transition probability matrix; otherwise, on the premise of ensuring that there are people in each air-conditioning supply area, generate probability distribution random numbers through the Monte Carlo method, and obtain the passenger number distribution in different air-conditioning supply areas according to the initial state and the transition probability matrix;

[0042] The specific content of S22 is as follows:

[0043] Assume that the number of air-conditioning supply air areas in each functional area is n, and the numbers of different air-conditioning supply air areas are the set I = {1, 2, …, n}. Then the transition probability matrix of the Markov chain is expressed as:

[0044]

[0045] Among them, each element p in the matrix ij represents the probability of transitioning from state i to state j, that is, it represents the probability that the people in the area are in the air-conditioning supply air area i at time t and in the air-conditioning supply air area j at time t + 1;

[0046] The transition probability matrix P t satisfies: all elements are non-negative and the sum of the elements in any row is 1.

[0047] Preferably, in S3, the optimization objective function includes the operating energy consumption of the air-conditioning system, the tracking performance of thermal and humidity comfort, and the tracking performance of regional air quality. The construction method is specifically as follows:

[0048] The operating energy consumption model of the air-conditioning system includes refrigeration energy consumption and fan energy consumption. The total energy consumption J1(t) is expressed as:

[0049] J1(t) = P c (t) + P f (t) (12)

[0050] Among them, P c (t) represents the refrigeration energy consumption, in kW; P f (t) represents the fan energy consumption, in kW;

[0051] The calculation formula for the refrigeration energy consumption P c (t) is specifically as follows:

[0052]

[0053] h ma (t) = (1 - dr(t))h ra (t) + dr(t)h oa (t) (14)

[0054] Among them, ρ a represents the air density, in kg / m 3 ; G sa (t) represents the supply air volume flow rate, in m 3 / s; h ma (t) represents the mixed air enthalpy value, in kJ / kg; h sa (t) represents the supply air enthalpy value, in kJ / kg; dr(t) represents the fresh air ratio; h ra(t) represents the return air enthalpy value, kJ / kg; h oa (t) represents the fresh air enthalpy value, kJ / kg; COP represents the refrigeration energy efficiency ratio;

[0055] The fan energy consumption P f (t) The calculation formula is as follows:

[0056] P f (t) = σG sa (t) 2 (15)

[0057] Among them, σ represents the penalty coefficient of the fan power function; G sa (t) represents the supply air volume flow rate, m 3 / s;

[0058] The thermal and humidity comfort tracking target J2(t) is expressed as:

[0059]

[0060] Among them, T in (t) represents the zone temperature, °C; T set (t) represents the zone temperature setpoint, °C; represents the zone relative humidity; represents the zone relative humidity setpoint;

[0061] Relative humidity is a function of temperature T in and humidity W in The calculation formula is as follows:

[0062]

[0063] Among them, P w,s represents the saturated water vapor partial pressure, kPa; P w represents the water vapor partial pressure, kPa; B represents the atmospheric pressure, with a value of 101.325 kPa;

[0064] The zone air quality tracking target J3 is expressed as:

[0065] J3 = (C in (t) - C min ) 2 (20)

[0066] Among them, C in (t) represents the zone carbon dioxide concentration, ppm; C min represents the lower limit of the carbon dioxide volume fraction, that is, the outdoor carbon dioxide volume fraction;

[0067] In S3, the specific method for clarifying the optimization objectives under different passenger distribution characteristics in the area is as follows:

[0068] When the area is in the occupied state, the upper thermal and humidity control layer takes the multi-objective function of thermal and humidity comfort tracking based on weight distribution and air-conditioning system operation energy consumption in the prediction time domain as the optimization objective, and the lower fresh air control layer takes the multi-objective function of area air quality and air-conditioning system operation energy consumption based on weight distribution in the prediction time domain as the optimization objective, specifically as follows:

[0069] Optimization objective of the thermal and humidity control layer in the occupied area It is expressed as:

[0070]

[0071] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω1 represents the weight coefficient of the system operation energy consumption target in the thermal and humidity control layer; J1(t) represents the system operation energy consumption target value at time t; ω2 represents the weight coefficient of the thermal and humidity comfort tracking target in the thermal and humidity control layer; J2(t) represents the thermal and humidity comfort tracking target value at time t;

[0072] Optimization objective of the fresh air control layer in the occupied area It is expressed as:

[0073]

[0074] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω3 represents the weight coefficient of the system operation energy consumption target in the fresh air control layer; J1(t) represents the system operation energy consumption target value at time t; ω4 represents the weight coefficient of the area air quality tracking target in the fresh air control layer; J3(t) represents the area air quality tracking target value at time t;

[0075] When the area is in the unoccupied state, the tracking of the area temperature, humidity and carbon dioxide concentration is not considered, and it only needs to be maintained within the constraints. Only the optimization objective of the air-conditioning system operation energy consumption is considered, specifically as follows:

[0076] The optimization objective of the thermal and humidity control layer in the unoccupied area, expressed as:

[0077]

[0078] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; J1(t) represents the total system operation energy consumption, kW; P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW;

[0079] The optimization objective of the fresh air control layer in the unoccupied area, expressed as:

[0080]

[0081] Among them, \(i\) represents the current control time, \(h\); \(L\) represents the prediction horizon length, \(h\); \(J_1(t)\) represents the total energy consumption of the system operation, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\);

[0082] When calculating the operation energy consumption of the air-conditioning system, the fresh air ratio in the heat and humidity control layer is a fixed value, and the remaining supply air state parameters are optimization variables. The supply air state parameters in the fresh air control layer are the optimization results of the heat and humidity control layer, and the fresh air ratio is an optimization variable.

[0083] Preferably, in the step S4, the parameter setting of the double-layer control model includes the parameters of each area, the control parameters of the model predictive control algorithm, and the initial parameters of the decision variables. The control parameters include the control horizon and the prediction horizon. Among them, the prediction horizon data includes outdoor meteorological data, the Markov transition probability matrix of each area, the number of passengers and occupancy in each area, the air-conditioning operation parameters, and the remaining heat disturbance data. The outdoor meteorological parameters include outdoor temperature, outdoor relative humidity, and solar radiation intensity. The remaining heat disturbance data includes the heat and humidity loads caused by infiltration, solar radiation, equipment, and lighting.

[0084] The double-layer control model includes a heat and humidity control layer and a fresh air control layer. The constraint conditions of the heat and humidity control layer include the regional temperature change constraint, the regional relative humidity change constraint, the supply air volume flow constraint, the supply air temperature constraint, and the supply air humidity constraint. The constraint conditions of the fresh air control layer include the regional carbon dioxide concentration constraint and the fresh air ratio constraint, which are specifically as follows:

[0085] The regional temperature change constraint is:

[0086]

[0087] Among them, represents the lower limit of the regional temperature, \(^{\circ}C\); represents the upper limit of the regional temperature, \(^{\circ}C\); \(T\) in represents the regional temperature, \(^{\circ}C\);

[0088] The regional relative humidity change constraint is:

[0089]

[0090] Among them, represents the lower limit of the regional relative humidity; represents the upper limit of the regional relative humidity; represents the regional relative humidity;

[0091] The supply air volume flow constraint is:

[0092]

[0093] Among them, represents the lower limit of the supply air volume flow rate, m 3 / s; represents the upper limit of the supply air volume flow rate, m 3 / s; G sa represents the supply air volume flow rate, m 3 / s;

[0094] The supply air temperature constraint is:

[0095]

[0096] Among them, represents the lower limit of the supply air temperature, °C; represents the upper limit of the supply air temperature, °C; T sa represents the supply air temperature, °C;

[0097] The supply air humidity constraint is:

[0098]

[0099] Among them, represents the lower limit of the supply air humidity, g / kg; represents the upper limit of the supply air humidity, g / kg; W sa represents the supply air humidity, g / kg;

[0100] The regional carbon dioxide concentration constraint:

[0101]

[0102] Among them, represents the lower limit of the regional carbon dioxide concentration, ppm; represents the upper limit of the regional carbon dioxide concentration, ppm; C in represents the regional carbon dioxide concentration, ppm;

[0103] The fresh air ratio constraint:

[0104] dr min ≤ dr ≤ dr max (31)

[0105] Among them, dr min represents the lower limit of the fresh air ratio; dr max represents the upper limit of the fresh air ratio; dr represents the fresh air ratio.

[0106] Preferably, the rolling optimization model of the double-layer control model in S5 is specifically as follows:

[0107] In the occupied state, the rolling optimization model of the thermal and humidity control layer is expressed as:

[0108]

[0109] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω1 represents the weight coefficient of the thermal and humidity comfort tracking target of the thermal and humidity control layer; J1(t) represents the thermal and humidity comfort tracking target value at time t; ω2 represents the weight coefficient of the system operation energy consumption target of the thermal and humidity control layer; J2(t) represents the system operation energy consumption target value at time t;

[0110] The rolling optimization model of the fresh air control layer is expressed as:

[0111]

[0112] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω3 represents the weight coefficient of the system operation energy consumption target of the fresh air control layer; J1(t) represents the system operation energy consumption target value at time t; ω4 represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; J3(t) represents the regional air quality tracking target value at time t;

[0113] In the unoccupied state, the rolling optimization model of the thermal and humidity control layer is expressed as:

[0114]

[0115] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW;

[0116] The rolling optimization model of the fresh air control layer is expressed as:

[0117]

[0118] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW.

[0119] The present invention has the following beneficial effects compared with the prior art:

[0120] The internal space of the terminal building is tall and has diverse functional areas, and there are significant differences in the passenger occupancy rates and environmental requirements of different areas. Starting from the spatial heterogeneity of the functional areas of the terminal building, the present invention deeply analyzes the multi-dimensional coupling relationship between the passenger flow density, the air supply parameters (air supply volume, air supply temperature, air supply humidity), and the fresh air ratio and the regional environmental parameters (temperature and humidity, CO2 concentration). Based on this, a differential control strategy is implemented for the occupied and unoccupied areas within the region, a two-layer control architecture including upper-layer heat and humidity control and lower-layer fresh air control is constructed, and a model predictive control algorithm is adopted. By accurately analyzing and dynamically adjusting the passenger distribution characteristics of different areas, the precise regulation of the air-conditioning operation parameters of each sub-region of the terminal building is realized. Compared with the traditional control strategy, the present invention effectively solves the problems of local pollutant accumulation and heat and humidity environment imbalance caused by uneven passenger flow distribution by introducing a dynamic zoning control mechanism. While ensuring the air quality and passenger comfort of each functional area of the terminal building, this method can significantly reduce the operating energy consumption of the air-conditioning system, providing a new technical path for the optimization of the terminal building air-conditioning system. Description of the Drawings

[0121] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0122] Figure 1 It is a control flow chart of a two-layer optimization control method for a terminal building air-conditioning system considering the spatio-temporality of passenger flow provided by the present invention.

[0123] Figure 2 It is a logic diagram of two-layer model predictive control provided by the present invention. Detailed Embodiments

[0124] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0125] Such as Figure 1As shown in the figure, a double-layer optimization control method for the terminal building air-conditioning system considering the spatio-temporal characteristics of passenger flow provided by the present invention. This control method is based on the passenger movement mathematical model, divides each functional area of the terminal building into an occupied area and an unoccupied area, and implements a differential control strategy for different air-conditioning supply areas; adopts a double-layer model predictive control algorithm, comprehensively considers the dynamic changes of passenger flow and the uncertainty of outdoor meteorological conditions, combines the real-time state information of the system, and through the collaborative optimization of the heat and humidity control layer and the fresh air control layer, rolls to solve the decision variables within the prediction time domain, including the optimal control sequences of supply air temperature, supply air humidity, supply air volume flow rate and fresh air ratio, and executes the corresponding control instructions.

[0126] The control method of the present invention specifically includes the following steps:

[0127] S1: Define the division principle of the terminal building air-conditioning control area, and construct a regional heat and humidity balance model and a regional air quality model considering the passenger occupancy rate.

[0128] As a preferred embodiment of the present invention, in this step, the division principle of the terminal building air-conditioning control area includes dividing the terminal building area into a check-in area, a security check area, and a waiting area according to the functional differences of different areas, and further dividing each of the three main functional areas into several sub-areas according to the air-conditioning supply area to formulate the air-conditioning control methods for different areas.

[0129] Specifically, the construction method of the regional heat and humidity balance model is as follows:

[0130] The regional heat and humidity balance model includes a regional heat balance model and a regional moisture balance model. The regional sensible heat load is mainly composed of internal heat sources (passengers, equipment, lighting) and external heat sources (envelope structure, solar radiation, infiltration). The change value of the regional air sensible heat is equal to the difference between the regional sensible heat load and the air-conditioning cooling capacity, and the regional heat balance model is obtained, which is expressed as:

[0131]

[0132] Among them, C p represents the specific heat capacity of regional air, kJ / (kg·K); ρ a represents the air density of the region, kg / m 3 ; V represents the regional environmental volume, m 3 ; T in represents the regional air temperature, °C; t represents the current moment, s; T oa represents the outdoor air temperature, °C; j represents the type of envelope structure; S represents the set of envelope structures in the region; A j represents the area of the envelope structure j, m 2 ; k j represents the heat transfer coefficient of the envelope structure j, W / (m 2·°C); G sa represents the supply air volume flow rate, m 3 / s; T sa represents the supply air temperature, °C; Q person represents the sensible heat dissipation of passengers, kW; Q other represents the heat generation of other heat sources, kW.

[0133] Q other = Q solar + Q infilt + Q device + Q light (2)

[0134]

[0135] Among them, Q solar is the heat gain from solar radiation, kW; Q infilt is the heat generated by air infiltration, kW; Q device is the heat dissipation of equipment, kW; Q light is the heat dissipation of lighting, kW; N represents the number of passengers in the area, person.

[0136] The latent heat load of the area is mainly composed of infiltration and fresh air moisture load, and the latent heat dissipation load of passengers. The change value of the latent heat of the air in the area is equal to the difference between the latent heat load of the area and the dehumidification amount of the air conditioner, and the area moisture balance model is obtained, which is expressed as:

[0137]

[0138]

[0139] M infilt = A inf Vρ a (W oa - W in ) (6)

[0140] Among them, the latent heat load of the area includes infiltration and fresh air moisture load, and the latent heat dissipation load of passengers; W in represents the air humidity in the area, g / kg; M person represents the latent heat dissipation load of passengers, g / s; M infilt represents the moisture load generated by air infiltration, g / s; W sa represents the supply air humidity, g / kg; N represents the number of passengers in the area, person; Δh vap represents the specific enthalpy of evaporation, 2257 J / g for water; A inf represents the infiltration air times, s -1 ; W oa represents the outdoor air humidity, g / kg.

[0141] The construction method of the regional air quality model is as follows:

[0142] There are various types of pollutants emitted in the terminal area, which are difficult to monitor and control. The amount of pollutants emitted by the human body is roughly proportional to the carbon dioxide concentration emitted by breathing. Therefore, the carbon dioxide concentration is used to measure the regional air quality. According to the mass balance equation, the regional carbon dioxide concentration calculation model is obtained:

[0143]

[0144] C ma = drC oa +(1 - dr)C in (8)

[0145] G in = G person N (9)

[0146] Among them, C in represents the regional carbon dioxide volume fraction; C ma represents the carbon dioxide volume fraction of the mixed air (return air and fresh air); C oa represents the outdoor carbon dioxide volume fraction, with a value of 400×10 -6 (i.e., 400 ppm); G sa represents the regional supply air volume flow rate, m 3 / s; V represents the regional environmental volume, m 3 ; A inf represents the infiltration air rate, s -1 ; dr represents the fresh air ratio, that is, the ratio of the fresh air volume to the supply air volume; G in represents the total rate of carbon dioxide generation by passengers in the region, m 3 / s; N represents the number of passengers in the region; G person represents the carbon dioxide generation rate of passengers, m 3 / (s·person).

[0147] The calculation formula for the carbon dioxide release rate per passenger is expressed as:

[0148]

[0149] Among them, G person represents the carbon dioxide generation rate per passenger, m 3 / (s·person); A D represents the DuBois surface area, m 2 , with an average of 1.8 m 2; M is the physical activity level in Met, which is set according to the activity types of passengers in the terminal; RQ represents the respiratory quotient, which is the molecular ratio of CO2 released and O2 absorbed during respiration. Generally, the average value is 0.83 under light physical labor.

[0150] S2: Construct a passenger movement model based on Markov chain and events, describe the distribution characteristics of passengers in the area through the transition probability matrix, and divide the occupied area and unoccupied area of each area in real time.

[0151] As a preferred embodiment of the present invention, the passenger movement model based on Markov chain and events is specifically as follows:

[0152] Constructing a passenger movement model based on the Markov transition probability matrix mainly includes the following steps:

[0153] (1) Taking the average value of the predicted number of passengers on the day as the standard, define an event mechanism based on the passenger flow.

[0154] (2) Obtain the number of passengers in different functional areas (check-in area, security check area, waiting area) under the current control time domain, and based on the defined event mechanism, obtain the passenger transfer probability matrix.

[0155] (3) Based on the passenger flow and transfer probability matrix under the current control time domain, calculate the passenger distribution in different air supply areas of each functional area according to the Markov movement model, and divide the occupied area and unoccupied area.

[0156] Use Markov chain to describe the distribution of passengers in the area. Regarding passengers as individual movement units, define the transition probability matrix under different event mechanisms for their movement methods, depict and simulate the displacement of passengers in the area, so as to generate the hourly passenger occupancy in sub-areas.

[0157] Assume that the number of air supply areas in each functional area is n, and the numbers of different air supply areas are in the set I = {1, 2,..., n}. Then the transition probability matrix of the Markov chain is expressed as:

[0158]

[0159] Among them, each element p in the matrix ij represents the probability of transitioning from state i to state j, that is, it represents the probability that the people in the area are in the air supply area i at time t and in the air supply area j at time t + 1.

[0160] The transition matrix P t satisfies the following properties:

[0161] 1) All elements are non-negative:

[0162] 2) The sum of elements in any row is 1:

[0163] Taking the average value of the predicted number of passengers on the same day as the standard, an event mechanism based on passenger flow is defined, expressed as:

[0164] If the number of passengers N(t) in the functional area of the current control horizon is less than the average number of passengers on the same day Then the transfer probability p of half of the air supply area for personnel ij Is set to zero (equivalent to the unoccupied area), and the remaining air supply areas generate probability distribution random numbers through the Monte Carlo method, and the passenger number distribution of each sub-area is obtained according to the initial state and the transition probability matrix.

[0165] Otherwise, on the premise of ensuring that there are people in each air supply area, probability distribution random numbers are generated through the Monte Carlo method, and the passenger number distribution of different air supply areas is obtained according to the initial state and the transition probability matrix.

[0166] S3: Based on the regional heat and moisture balance model and the regional air quality model in S1, construct the optimization objective function of the double-layer control model, and clarify the optimization objectives of the region under different passenger distribution characteristics.

[0167] As a relatively optimal embodiment of the present invention, the construction of the optimization objective function is specifically as follows:

[0168] The objective function includes the operating energy consumption of the air conditioning system, the tracking performance of thermal and moisture comfort, and the tracking performance of regional air quality.

[0169] The operating energy consumption model of the air conditioning system includes refrigeration energy consumption and fan energy consumption, and the total energy consumption J1(t) is expressed as:

[0170] J1(t) = P c (t) + P f (t) (12)

[0171] Among them, P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW.

[0172] The refrigeration energy consumption P c (t) calculation formula is specifically as follows:

[0173]

[0174] h ma (t) = (1 - dr(t))h ra (t) + dr(t)h oa (t) (14)

[0175] Among them, ρa represents the air density, kg / m 3 ; G sa (t) represents the supply air volume flow rate, m 3 / s; h ma (t) represents the enthalpy value of the mixed air (supply air and return air), kJ / kg; h sa (t) represents the enthalpy value of the supply air, kJ / kg; dr(t) represents the fresh air ratio; h ra (t) represents the enthalpy value of the return air, kJ / kg; h oa (t) represents the enthalpy value of the fresh air, kJ / kg.

[0176] The fan energy consumption P f (t) calculation formula is as follows:

[0177] P f (t) = σG sa (t) 2 (15)

[0178] where, σ represents the penalty coefficient of the fan power function; G sa (t) represents the supply air volume flow rate, m 3 / s

[0179] The thermal and humidity comfort tracking target J2(t) is expressed as:

[0180]

[0181] where, T in (t) represents the zone temperature, °C; T set (t) represents the zone temperature setpoint, °C; represents the zone relative humidity; represents the zone relative humidity setpoint.

[0182] Relative humidity is a function of temperature T in and humidity W in The calculation formula is as follows:

[0183]

[0184] where, P w,s represents the saturated water vapor partial pressure, kPa; P w represents the water vapor partial pressure, kPa; B represents the atmospheric pressure, with a value of 101.325 kPa; W is the humidity of the moist air, g / kg; represents the relative humidity.

[0185] The zone air quality tracking target J3 is expressed as:

[0186] J3 = (Cin (t)-C min ) 2 (20)

[0187] Among them, C in (t) represents the regional carbon dioxide concentration, ppm; C min represents the lower limit of the carbon dioxide volume fraction, that is, the outdoor carbon dioxide volume fraction.

[0188] In S3, clarify the optimization objectives of the area under different passenger distribution characteristics, specifically as follows:

[0189] When the area is in the occupied state, the upper thermal and humidity control layer takes the multi-objective function of thermal and humidity comfort tracking based on weight distribution and air-conditioning system operation energy consumption in the prediction time domain as the optimization objective, and the lower fresh air control layer takes the multi-objective function of regional air quality and air-conditioning system operation energy consumption based on weight distribution in the prediction time domain as the optimization objective, specifically as follows:

[0190] The optimization objective of the thermal and humidity control layer in the occupied area is expressed as:

[0191]

[0192] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω1 represents the weight coefficient of the system operation energy consumption target in the thermal and humidity control layer; J1(t) represents the system operation energy consumption target value at time t; ω2 represents the weight coefficient of the thermal and humidity comfort tracking target in the thermal and humidity control layer; J2(t) represents the thermal and humidity comfort tracking target value at time t;

[0193] The optimization objective of the fresh air control layer in the occupied area is expressed as:

[0194]

[0195] Among them, i represents the current control moment, h; L represents the prediction time domain length, h; ω3 represents the weight coefficient of the system operation energy consumption target in the fresh air control layer; J1(t) represents the system operation energy consumption target value at time t; ω4 represents the weight coefficient of the regional air quality tracking target in the fresh air control layer; J3(t) represents the regional air quality tracking target value at time t;

[0196] When the area is in the unoccupied state, do not consider tracking the temperature, humidity and carbon dioxide concentration of the area, just maintain it within the constraints, and only consider the optimization objective of the air-conditioning system operation energy consumption, specifically as follows:

[0197] The optimization objective of the thermal and humidity control layer in the unoccupied area is expressed as:

[0198]

[0199] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(J_1(t)\) represents the total energy consumption of the system operation, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

[0200] The optimization objective of the fresh air control layer in the unoccupied area is expressed as:

[0201]

[0202] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(J_1(t)\) represents the total energy consumption of the system operation, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

[0203] When calculating the energy consumption of the air conditioning system operation, the fresh air ratio of the heat and moisture control layer is a fixed value, and the remaining supply air state parameters are optimization variables. The supply air state parameters (supply air temperature, supply air humidity, supply air volume flow rate) of the fresh air control layer are the optimization results of the heat and moisture control layer, and the fresh air ratio is an optimization variable.

[0204] S4: Set the parameters of the double-layer control model in S3, clarify the constraint conditions, and obtain the data of the prediction time domain.

[0205] As a relatively optimal embodiment of the present invention, in this step, clarify the constraint conditions of the double-layer control, specifically as follows:

[0206] The double-layer control includes a heat and moisture control layer and a fresh air control layer. The constraint conditions of the heat and moisture control layer include regional temperature change constraint, regional relative humidity change constraint, supply air volume flow rate constraint, supply air temperature constraint, and supply air humidity constraint.

[0207] The regional temperature change constraint is:

[0208]

[0209] Among them, represents the lower limit of the regional temperature, \(^{\circ}C\); represents the upper limit of the regional temperature, \(^{\circ}C\).

[0210] The regional relative humidity change constraint is:

[0211]

[0212] Among them, represents the lower limit of the regional relative humidity; represents the upper limit of the regional relative humidity.

[0213] The supply air volume flow rate constraint is:

[0214]

[0215] Among them, represents the lower limit of the supply air volume flow rate, m 3 / s; represents the upper limit of the supply air volume flow rate, m 3 / s.

[0216] The supply air temperature constraint is:

[0217]

[0218] Among them, represents the lower limit of the supply air temperature, °C; represents the upper limit of the supply air temperature, °C.

[0219] The supply air humidity constraint is:

[0220]

[0221] Among them, represents the lower limit of the supply air humidity, g / kg; represents the upper limit of the supply air humidity, g / kg.

[0222] The constraint conditions of the fresh air control layer include the regional carbon dioxide concentration constraint and the fresh air ratio constraint.

[0223] Regional carbon dioxide concentration constraint:

[0224]

[0225] Among them, represents the lower limit of the regional carbon dioxide concentration, ppm; represents the upper limit of the regional carbon dioxide concentration, ppm.

[0226] Fresh air ratio constraint:

[0227] dr min ≤ dr ≤ dr max (31)

[0228] Among them, dr min represents the lower limit of the fresh air ratio; dr max represents the upper limit of the fresh air ratio.

[0229] S5: Based on the passenger movement model in S2 and the results of S3, construct a two-layer control model for the occupied area and the unoccupied area; the upper thermal and humidity control layer optimizes the supply air temperature, supply air humidity, and supply air volume flow rate according to the passenger density, outdoor weather conditions, and system status information, and the lower fresh air control layer optimizes the fresh air ratio according to the regional air quality requirements.

[0230] As a preferred embodiment of the present invention, in this step, the rolling optimization model for dual-layer control is specifically as follows:

[0231] In the occupied state, the rolling optimization model of the heat and moisture control layer can be expressed as:

[0232]

[0233] s.t Equations (1)-(6), (12)-(19), (25)-(29)

[0234] Where, i represents the current control moment, h; L represents the prediction time domain length, h; ω1 represents the weight coefficient of the heat and moisture comfort tracking target of the heat and moisture control layer; J1(t) represents the heat and moisture comfort tracking target value at time t; ω2 represents the weight coefficient of the system operation energy consumption target of the heat and moisture control layer; J2(t) represents the system operation energy consumption target value at time t.

[0235] The rolling optimization model of the fresh air control layer can be expressed as:

[0236]

[0237] s.t Equations (7)-(10), (12)-(15), (20), (30)-(31)

[0238] Where, i represents the current control moment, h; L represents the prediction time domain length, h; ω3 represents the weight coefficient of the system operation energy consumption target of the fresh air control layer; J1(t) represents the system operation energy consumption target value at time t; ω4 represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; J3(t) represents the regional air quality tracking target value at time t.

[0239] In the unoccupied state, the rolling optimization model of the heat and moisture control layer can be expressed as:

[0240]

[0241] s.t Equations (1)-(6), (12)-(15), (17)-(19), (25)-(29)

[0242] Where, i represents the current control moment, h; L represents the prediction time domain length, h; P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW.

[0243] The rolling optimization model of the fresh air control layer can be expressed as:

[0244]

[0245] s.t. Equations (7)-(10), (12)-(15), (30)-(31).

[0246] S6: Execute the control instructions within the control time domain, update the state information of the two-layer control model in S5, and feedback the updated state information to the next control time domain. Repeat steps S2 - S5 and perform rolling optimization until the scheduling period ends.

[0247] The method and effect of the present invention will be further described below through embodiments.

[0248] Embodiment

[0249] In the prior art, the air-conditioning control method mainly takes the regional temperature as the optimization control target to reduce the energy consumption of the system. However, it usually ignores the influence of humidity and fresh air, and it is difficult to take into account both the thermal and humidity comfort and the air quality target at the same time, which may lead to problems such as thermal and humidity imbalance and regional pollutant accumulation.

[0250] In this embodiment, one day is used as the scheduling period, that is, the scheduling period is 24 hours, the control time domain is set to 1 hour, and the prediction time domain is 6 hours: data collection and optimization solution are performed every 1 hour. Considering the uncertainty of changes such as outdoor weather and passenger occupancy rate, the model predictive control algorithm is used to predict the system state changes in the next 6h, solve the optimal control sequence, and continuously perform rolling optimization over time until the scheduling period ends.

[0251] The control flow of the control method in this embodiment is as Figure 1 shown, and the specific implementation steps are as follows:

[0252] S1: Define the division principle of the air-conditioning control area in the terminal building, and construct a regional thermal and humidity balance model and a regional air quality model considering passenger occupancy rate;

[0253] S11: Division of the air-conditioning control area in the terminal building

[0254] Specifically, according to the functional differences of different areas in the terminal building, the area is divided into the check-in area, the security check area, and the waiting area, and several sub-areas are further divided according to the air-conditioning supply area within the three main functional areas. Define the air-conditioning control area in the terminal building to perform temperature, humidity, and air quality control in different areas respectively.

[0255] S12: Regional thermal and humidity balance model

[0256] The regional heat and moisture balance model includes a regional heat balance model and a regional moisture balance model. The regional sensible heat load is mainly composed of internal heat sources (passengers, equipment, lighting) and external heat sources (envelope structure, solar radiation, infiltration). The change value of the sensible heat of the regional air is equal to the difference between the regional sensible heat load and the air-conditioning cooling capacity, obtaining the regional heat balance model, expressed as:

[0257]

[0258] Among them, C p represents the specific heat capacity of the regional air, with a value of 1.005 kJ / (kg·K); ρ a represents the density of the regional air, with a value of 1.205 kg / m 3 ; V represents the volume of the regional environment, m 3 ; T in represents the temperature of the regional air, °C; t represents the current moment, s; T oa represents the outdoor air temperature, °C; j represents the type of envelope structure; S represents the set of envelope structures in the region; A j represents the area of the envelope structure j, m 2 ; k j represents the heat transfer coefficient of the envelope structure j, W / (m 2 ·°C); G sa represents the supply air volume flow rate, m 3 / s; T sa represents the supply air temperature, °C; Q person represents the sensible heat dissipation of passengers, kW; Q other represents the heat generation of other heat sources, kW.

[0259] Q other =Q solar +Q infilt +Q device +Q light (2)

[0260]

[0261] Among them, Q solar is the heat gain from solar radiation, kW; Q infilt is the heat generated by air infiltration, kW; Q device is the heat dissipation of equipment, kW; Q light is the heat dissipation of lighting, kW; N represents the number of passengers in the region, person.

[0262] The regional latent heat load is mainly composed of infiltration and fresh air moisture load, and the latent heat dissipation load of passengers. The change value of the latent heat of the regional air is equal to the difference between the regional latent heat load and the air-conditioning dehumidification capacity, obtaining the regional moisture balance model, expressed as:

[0263]

[0264] M infilt = A inf Vρ a (W oa - W in ) (6)

[0265] Among them, the regional latent heat load includes infiltration and fresh air moisture load, and the latent heat dissipation load of passengers; W in represents the regional air humidity, g / kg; M person represents the latent heat dissipation load of passengers, g / s; M infilt represents the moisture load generated by air infiltration, g / s; W sa represents the supply air humidity, g / kg; N represents the number of passengers in the area, person; Δh vap represents the specific enthalpy of evaporation, 2257 J / g for water; A inf represents the infiltration air rate, s -1 ; W oa represents the outdoor air humidity, g / kg.

[0266] S13: Regional air quality model

[0267] Specifically, there are various types of pollutants emitted in the terminal area, which are difficult to monitor and control. The amount of pollutants emitted by the human body is roughly proportional to the carbon dioxide concentration emitted by breathing. Therefore, the carbon dioxide concentration is used to measure the regional air quality. According to the mass balance equation, the regional carbon dioxide concentration calculation model is obtained:

[0268]

[0269] C ma = drC oa + (1 - dr)C in (8)

[0270] G in = G person N (9)

[0271] Among them, C in represents the regional carbon dioxide volume fraction; C ma represents the carbon dioxide volume fraction of the mixed air (return air and fresh air); C oa represents the outdoor carbon dioxide volume fraction, with a value of 400×10 -6 (i.e., 400 ppm); G sa represents the regional supply air volume flow rate, m 3 / s; V represents the regional environmental volume, m 3 ; A inf represents the infiltration air rate, s -1; dr represents the fresh air ratio; G in represents the total rate of carbon dioxide generation by passengers in the area, m 3 / s; N represents the number of passengers in the area; G person represents the rate of carbon dioxide generation by passengers, m 3 / (s·person).

[0272] Furthermore, the calculation formula for the per capita carbon dioxide release rate of passengers is expressed as:

[0273]

[0274] Among them, G person represents the per capita carbon dioxide generation rate of passengers, m 3 / (s·person); A D represents the DuBois surface area, m 2 , with an average of 1.8 m 2 ; M is the physical activity level in Met, which is set according to the activity types of passengers in the terminal; RQ represents the respiratory quotient, which is the molecular ratio of CO2 released and O2 absorbed during respiration. Generally, the average value under light physical labor is 0.83.

[0275] According to the relevant requirements of GB / T 18883-2002 "Indoor Air Quality Standard", the upper limit of the volume fraction of carbon dioxide in the terminal is set to 1000 ppm in this embodiment.

[0276] S2: Construct a passenger movement model based on Markov chain and events, describe the distribution characteristics of passengers in the area through the transition probability matrix, and divide the occupied area and unoccupied area of each area in real time;

[0277] Specifically, constructing a passenger movement model based on the Markov transition probability matrix mainly includes the following steps:

[0278] (1) Taking the average value of the predicted number of passengers on the same day as the standard, define an event mechanism based on passenger flow

[0279] (2) Obtain the number of passengers in different functional areas (check-in area, security check area, waiting area) under the current control time domain, and based on the defined event mechanism, obtain the passenger transition probability matrix

[0280] (3) Based on the passenger flow and transition probability matrix under the current control time domain, calculate the passenger distribution in different air-conditioning supply areas in each functional area according to the Markov movement model, and divide the occupied area and unoccupied area.

[0281] A Markov chain is used to describe the distribution of passengers in the area. Taking passengers as individual moving units, by defining the transition probability matrix under different event mechanisms for their moving modes, the displacement of passengers in the area is characterized and simulated to generate the hourly passenger occupancy in sub-areas.

[0282] Furthermore, assuming that the number of air-conditioning supply areas in each functional area is n, and the numbers of different air-conditioning supply areas are in the set I = {1, 2, …, n}, then the transition probability matrix of the Markov chain is expressed as:

[0283]

[0284] Among them, each element p in the matrix ij represents the probability of transitioning from state i to state j, that is, it represents the probability that the people in the area are in the air-conditioning supply area i at time t and in the air-conditioning supply area j at time t + 1.

[0285] The transition matrix P t satisfies the following properties:

[0286] 1) All elements are non-negative:

[0287] 2) The sum of the elements in any row is 1:

[0288] Furthermore, taking the average value of the predicted number of passengers on the same day as the standard, an event mechanism based on the passenger flow is defined to obtain the transition probability matrix at different times, which is expressed as:

[0289] If the number of passengers N(t) in the functional area of the current control time domain is less than the average number of passengers on the same day then the transition probability p of half of the air-conditioning supply areas for people ij is set to zero (equivalent to the unoccupied area), and the probability distributions of random numbers for the remaining air-conditioning supply areas are generated by the Monte Carlo method, and the passenger number distributions in each sub-area are obtained according to the initial state and the transition probability matrix.

[0290] Otherwise, on the premise of ensuring that there are people in each air-conditioning supply area, the probability distributions of random numbers are generated by the Monte Carlo method, and the passenger number distributions in different air-conditioning supply areas are obtained according to the initial state and the transition probability matrix.

[0291] S3: Based on the regional heat and moisture balance model and the air quality model, construct the optimization objective function of the double-layer control model, including the tracking of heat and moisture comfort, the energy consumption of the air-conditioning system, and the tracking of regional air quality, and clarify the optimization objectives of the area under different passenger distribution characteristics.

[0292] S31: Construct the optimization objective function

[0293] Specifically, the objective function includes the operating energy consumption of the air conditioning system, the tracking performance of thermal and humidity comfort, and the tracking performance of regional air quality. The operating energy consumption model of the air conditioning system includes the refrigeration energy consumption and the fan energy consumption. The total energy consumption J1(t) is expressed as:

[0294] J1(t) = P c (t) + P f (t) (12)

[0295] Where, P c (t) represents the refrigeration energy consumption, in kW; P f (t) represents the fan energy consumption, in kW.

[0296] The calculation formula for the refrigeration energy consumption P c (t) is specifically as follows:

[0297]

[0298] h ma (t) = (1 - dr(t))h ra (t) + dr(t)h oa (t) (14)

[0299] Where, ρ a represents the air density, in kg / m 3 ; G sa (t) represents the supply air volume flow rate, in m 3 / s; h ma (t) represents the enthalpy value of the mixed air (supply air and return air), in kJ / kg; h sa (t) represents the enthalpy value of the supply air, in kJ / kg; dr(t) represents the fresh air ratio; h ra (t) represents the enthalpy value of the return air, in kJ / kg; h oa (t) represents the enthalpy value of the fresh air, in kJ / kg; COP represents the refrigeration energy efficiency ratio.

[0300] The calculation formula for the fan energy consumption P f (t) is specifically as follows:

[0301] P f (t) = σG sa (t) 2 (15)

[0302] Where, σ represents the penalty coefficient of the fan power function; G sa (t) represents the supply air volume flow rate, in m 3 / s.

[0303] Specifically, the thermal and humidity comfort tracking target J2(t) is expressed as:

[0304]

[0305] Among them, T in (t) represents the regional temperature, °C; T set (t) represents the regional temperature set value, °C; represents the regional relative humidity; represents the regional relative humidity set value.

[0306] Furthermore, the relative humidity is a function of temperature T in and humidity W in The calculation formula is as follows:

[0307]

[0308] Among them, P w,s represents the saturated water vapor partial pressure, kPa; P w represents the water vapor partial pressure, kPa; B represents the atmospheric pressure, and the value is 101.325 kPa;

[0309] Specifically, the regional air quality tracking target J3 is expressed as:

[0310] J3 = (C in (t) - C min ) 2 (20)

[0311] Among them, C in (t) represents the regional carbon dioxide concentration, ppm; C min represents the lower limit of the carbon dioxide volume fraction, that is, the outdoor carbon dioxide volume fraction.

[0312] S31: Define the optimization objectives for different passenger distribution characteristics in the area

[0313] Specifically, define the optimization objectives for different passenger distribution characteristics in the area, and implement different control strategies for the occupied area and the unoccupied area. When the area is in the occupied state, the upper thermal and humidity control layer takes the multi-objective function of thermal and humidity comfort tracking based on weight distribution and the operating energy consumption of the air conditioning system within the prediction time domain as the optimization objective, and the lower fresh air control layer takes the multi-objective function of regional air quality and the operating energy consumption of the air conditioning system based on weight distribution within the prediction time domain as the optimization objective. When the area is in the unoccupied state, do not consider tracking the temperature, humidity and carbon dioxide concentration of the area, and only maintain it within a reasonable constraint range, and only consider the optimization objective of the operating energy consumption of the air conditioning system.

[0314] Furthermore, the double-layer optimization objectives for the occupied area and the unoccupied area are as follows:

[0315] The optimization objective of the thermal and humidity control layer in the occupied area is expressed as:

[0316]

[0317] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(\omega_1\) represents the weight coefficient of the operation energy consumption target of the heat and moisture control layer system; \(J_1(t)\) represents the operation energy consumption target value at time \(t\); \(\omega_2\) represents the weight coefficient of the heat and moisture comfort tracking target of the heat and moisture control layer; \(J_2(t)\) represents the heat and moisture comfort tracking target value at time \(t\).

[0318] The optimization objective of the fresh air control layer in the occupied area is expressed as:

[0319]

[0320] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(\omega_3\) represents the weight coefficient of the operation energy consumption target of the fresh air control layer system; \(J_1(t)\) represents the operation energy consumption target value at time \(t\); \(\omega_4\) represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; \(J_3(t)\) represents the regional air quality tracking target value at time \(t\).

[0321] The optimization objective of the heat and moisture control layer in the unoccupied area is expressed as:

[0322]

[0323] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(J_1(t)\) represents the total operation energy consumption of the system, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

[0324] The optimization objective of the fresh air control layer in the unoccupied area is expressed as:

[0325]

[0326] Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(J_1(t)\) represents the total operation energy consumption of the system, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

[0327] It should be noted that when calculating the operation energy consumption of the air conditioning system, the fresh air ratio of the heat and moisture control layer is a fixed value, and the remaining supply air state parameters are optimization variables. The supply air state parameters (supply air temperature, supply air humidity, supply air volume flow rate) of the fresh air control layer are the optimization results of the heat and moisture control layer, and the fresh air ratio is an optimization variable.

[0328] S4: Set the parameters of the double-layer control model, clarify the constraint conditions, and obtain the data of the prediction time domain.

[0329] Specifically, the parameter setting of the double-layer control model includes the parameters of each area, the control parameters (control time domain, prediction time domain) of the model predictive control algorithm, and the initial parameters of the decision variables. The prediction time domain data includes outdoor meteorological data, the Markov transition probability matrix of each area, the number of passengers and occupancy in each area, air-conditioning operation parameters, and other heat disturbance data.

[0330] Among them, the outdoor meteorological parameters include outdoor temperature, outdoor relative humidity, and solar radiation intensity, and the other heat disturbance data includes the heat and moisture loads caused by infiltration, solar radiation, equipment, and lighting.

[0331] Furthermore, the constraint conditions of the double-layer control are clarified as follows:

[0332] The double-layer control includes a heat and moisture control layer and a fresh air control layer. The constraint conditions of the heat and moisture control layer include the regional temperature change constraint, the regional relative humidity change constraint, the supply air volume flow constraint, the supply air temperature constraint, and the supply air humidity constraint.

[0333] The regional temperature change constraint is:

[0334]

[0335] Among them, represents the lower limit of the regional temperature, °C; represents the upper limit of the regional temperature, °C.

[0336] The regional relative humidity change constraint is:

[0337]

[0338] Among them, represents the lower limit of the regional relative humidity; represents the upper limit of the regional relative humidity.

[0339] The supply air volume flow constraint is:

[0340]

[0341] Among them, represents the lower limit of the supply air volume flow, m 3 / s; represents the upper limit of the supply air volume flow, m 3 / s.

[0342] The supply air temperature constraint is:

[0343]

[0344] Among them, represents the lower limit of the supply air temperature, °C; Represents the upper limit of the supply air temperature, °C.

[0345] The supply air humidity constraint is:

[0346]

[0347] Wherein, Represents the lower limit of the supply air humidity, g / kg; Represents the upper limit of the supply air humidity, g / kg.

[0348] The constraint conditions of the fresh air control layer include the regional carbon dioxide concentration constraint and the fresh air ratio constraint.

[0349] Regional carbon dioxide concentration constraint:

[0350]

[0351] Wherein, Represents the lower limit of the regional carbon dioxide concentration, ppm; Represents the upper limit of the regional carbon dioxide concentration, ppm.

[0352] Fresh air ratio constraint:

[0353] dr min ≤ dr ≤ dr max (31)

[0354] Wherein, dr min Represents the lower limit of the fresh air ratio; dr max Represents the upper limit of the fresh air ratio.

[0355] S5: Based on the passenger movement model, a two-layer control model for the occupied area and the unoccupied area is constructed. The upper-layer heat and humidity control layer optimizes the supply air temperature, supply air humidity, and supply air volume flow according to the passenger density, outdoor weather conditions, and system status information. The lower-layer fresh air control layer optimizes the fresh air ratio according to the regional air quality requirements.

[0356] The logic diagram of the two-layer model predictive control is as Figure 2 shown. The rolling horizon optimization strategy is adopted to execute the corresponding control strategies for the occupied area and the unoccupied area, as follows:

[0357] S51: Optimization model under the occupied state

[0358] Under the occupied state, the rolling optimization model of the heat and humidity control layer can be expressed as:

[0359]

[0360] s.t Equations (1)-(6), (12)-(19), (25)-(29)

[0361] Among them, \(i\) represents the current control time, \(h\); \(L\) represents the prediction time domain length, \(h\); \(\omega_1\) represents the weight coefficient of the thermal and humidity comfort tracking target of the thermal and humidity control layer; \(J_1(t)\) represents the thermal and humidity comfort tracking target value at time \(t\); \(\omega_2\) represents the weight coefficient of the system operation energy consumption target of the thermal and humidity control layer; \(J_2(t)\) represents the system operation energy consumption target value at time \(t\).

[0362] The rolling optimization model of the fresh air control layer can be expressed as:

[0363]

[0364] s.t Equation (7) - Equation (10), Equation (12) - Equation (15), Equation (20), Equation (30) - Equation (31)

[0365] Among them, \(i\) represents the current control time, \(h\); \(L\) represents the prediction time domain length, \(h\); \(\omega_3\) represents the weight coefficient of the system operation energy consumption target of the fresh air control layer; \(J_1(t)\) represents the system operation energy consumption target value at time \(t\); \(\omega_4\) represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; \(J_3(t)\) represents the regional air quality tracking target value at time \(t\).

[0366] S52: Optimization model in the unoccupied state

[0367] In the unoccupied state, the rolling optimization model of the thermal and humidity control layer can be expressed as:

[0368]

[0369] s.t Equation (1) - Equation (6), Equation (12) - Equation (15), Equation (17) - Equation (19), Equation (25) - Equation (29)

[0370] Among them, \(i\) represents the current control time, \(h\); \(L\) represents the prediction time domain length, \(h\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

[0371] The rolling optimization model of the fresh air control layer can be expressed as:

[0372]

[0373] s.t Equation (7) - Equation (10), Equation (12) - Equation (15), Equation (30) - Equation (31)

[0374] S6: Execute the control instructions within the control time domain, update the state information of the control model, and feedback the updated state information to the next control time domain. Repeat steps S2 - S5 and perform rolling optimization until the scheduling period ends.

[0375] Specifically, the control instructions include the supply air temperature, supply air humidity, supply air volume flow rate, and fresh air ratio within a 1-hour control time domain. The states of the control system include the zone temperature, zone humidity, carbon dioxide concentration, and number of passengers within a 1-hour control time domain, which are fed back to the next control moment as initial values.

[0376] Furthermore, a two-layer rolling optimization strategy is adopted to continuously adjust the control parameters according to the current operating state and external environmental changes. At each control moment, the system optimizes through a model predictive control algorithm based on the latest meteorological data, number of passengers, zone temperature and humidity, and air quality data, obtains the air conditioning operating parameters within the prediction time domain, and executes the corresponding control actions.

[0377] Based on the passenger movement mathematical model, this invention implements differential control strategies for areas such as the check-in area, security check area, and waiting area in the terminal building according to the division of occupied areas and unoccupied areas. Adopting a two-layer model predictive control algorithm, comprehensively considering the uncertainty of passenger flow dynamic changes and outdoor meteorological conditions, combining with the real-time state information of the system, through the collaborative optimization of the heat and moisture control layer and the fresh air control layer, the decision variables within the prediction time domain are solved by rolling, including the optimal control sequences of supply air temperature, supply air humidity, supply air volume flow rate, and fresh air ratio, and the corresponding control instructions are executed.

[0378] This invention divides the functional areas of the terminal building and constructs a regional heat and moisture balance model and an air quality model considering passenger occupancy. Based on the Markov chain and event-based passenger movement model, the transfer probability matrix is used to describe the flow of passengers in each area, divide the occupied areas and unoccupied areas, and implement differential control strategies for areas with different occupancy states. A multi-objective function with the optimization objectives of heat and moisture comfort tracking, air conditioning system operation energy consumption, and regional air quality tracking is constructed, and rolling optimization is carried out in combination with a two-layer control architecture. According to the passenger density, outdoor meteorological conditions, and system state information, the upper heat and moisture control layer optimizes the supply air temperature, supply air humidity, and supply air volume flow rate, and the lower fresh air control layer optimizes the fresh air ratio. Through the collaborative optimization of the heat and moisture control layer and the fresh air control layer, the air conditioning operating parameters of each area are dynamically adjusted to achieve precise control of the terminal building area environment. Compared with traditional control strategies, this invention effectively solves the problems of local pollutant accumulation and heat and moisture environment imbalance caused by uneven passenger flow distribution by introducing a dynamic zoning control mechanism, significantly reduces the operation energy consumption of the air conditioning system, achieves the balance of energy efficiency and comfort, and provides a new technical path for the optimization of the terminal building air conditioning system.

[0379] The embodiments described above are only a preferred solution of the present invention, but they are not intended to limit the present invention. Those of ordinary skill in the relevant technical field can still make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all technical solutions obtained by means of equivalent replacement or equivalent transformation fall within the protection scope of the present invention.

Claims

1. A double-layer optimal control method for the terminal air-conditioning system considering the spatio-temporal characteristics of passenger flow, characterized in that, Based on the mathematical model of passenger movement, the air supply areas in each functional area of the terminal are divided into occupied areas and unoccupied areas, and different control strategies are implemented for different air supply areas. The double-layer model predictive control algorithm is adopted, comprehensively considering the uncertainty of passenger dynamic changes and outdoor meteorological conditions, combining the real-time state information of the system, and through the collaborative optimization of the heat and humidity control layer and the fresh air control layer, the decision variables within the prediction horizon are solved iteratively, including the optimal control sequences of supply air temperature, supply air humidity, supply air volume flow rate, and fresh air ratio, and the corresponding control instructions are executed.

2. The double-layer optimization control method for the terminal building air-conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 1, characterized in that, The specific control method is as follows: S1: Define the division principle of the air-conditioning control area in the terminal, and construct a regional heat and humidity balance model and a regional air quality model considering passenger occupancy. S2: Construct a passenger movement model based on Markov chain and events, describe the distribution characteristics of passengers in the area through the transition probability matrix, and divide the occupied areas and unoccupied areas of each area in real time. S3: Based on the regional heat and humidity balance model and the regional air quality model in S1, construct the optimization objective function of the double-layer control model, and clarify the optimization objectives of the area under different passenger distribution characteristics. S4: Set the parameters of the double-layer control model in S3, clarify the constraint conditions, and obtain the data within the prediction horizon. S5: Based on the passenger movement model in S2 and the results of S3, construct a double-layer control model for the occupied area and the unoccupied area. The upper heat and humidity control layer optimizes the supply air temperature, supply air humidity, and supply air volume flow rate according to the passenger density, outdoor weather conditions, and system state information, and the lower fresh air control layer optimizes the fresh air ratio according to the regional air quality requirements. S6: Execute the control instructions within the control horizon, update the state information of the double-layer control model in S5, and feedback the updated state information to the next control horizon. Repeat steps S2 to S5 and perform iterative optimization until the scheduling period ends.

3. A double-layer optimization control method for a terminal air conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, wherein, In S1, the division principle of the air-conditioning control area in the terminal includes dividing the terminal area into the check-in area, the security check area, and the waiting area according to the functional differences of different areas, and further dividing each of the three main functional areas into several sub-areas according to the air supply area to formulate the air-conditioning control methods for different areas.

4. A double-layer optimization control method for an airport terminal air conditioning system considering the spatio-temporal characteristics of passenger flow, as claimed in claim 2, wherein In S1, the regional heat and humidity balance model includes a regional heat balance model and a regional moisture balance model. The construction method is as follows: For the regional heat balance model, it is calculated by the change value of the sensible heat of the regional air being equal to the difference between the regional sensible heat load and the air-conditioning refrigeration capacity, expressed as: Among them, the regional sensible heat load includes internal heat sources and external heat sources. The internal heat sources include passengers, equipment, and lighting, and the external heat sources include the building envelope, solar radiation, and infiltration; C p represents the specific heat capacity of the regional air, kJ / (kg·K); ρ a represents the density of the regional air, kg / m 3 ; V represents the volume of the regional space, m 3 ; T in represents the temperature of the regional air, °C; t represents the current time, s; T oa represents the temperature of the outdoor air, °C; j represents the type of the building envelope of the region; S represents the set of the building envelopes of the region; A j represents the area of the building envelope j, m 2 ; k j represents the heat transfer coefficient of the building envelope j, W / (m 2 ·°C); G sa represents the supply air volume flow rate, m 3 / s; T sa represents the supply air temperature, °C; Q person represents the sensible heat dissipation of passengers, kW; Q other represents the heat generation of the remaining heat sources, kW; Q other = Q solar + Q infilt + Q device + Q light (2) Among them, Q solar is the solar radiation heat gain, in kW; Q infilt is the heat transfer due to air infiltration, in kW; Q device is the heat dissipation from equipment, in kW; Q light is the heat dissipation from lighting, in kW; Q person represents the sensible heat dissipation of passengers, in kW; N represents the number of passengers in the area, in persons; For the regional moisture balance model, it is calculated by the change value of the latent heat of the regional air being equal to the difference between the regional latent heat load and the air-conditioning dehumidification capacity, expressed as: M infilt = A inf Vρ a (W oa - W in )(6) Among them, the regional latent heat load includes infiltration and fresh air moisture load, and the latent heat dissipation load of passengers; W in represents the regional air humidity, g / kg; M person represents the latent heat dissipation load of passengers, g / s; M infilt represents the moisture load generated by air infiltration, g / s; W sa represents the supply air humidity, g / kg; N represents the number of passengers in the area, person; Δh vap represents the specific enthalpy of evaporation, 2257 J / g for water; A inf represents the infiltration air rate, s -1 ; W oa represents the outdoor air humidity, g / kg.

5. The double-layer optimization control method for the terminal building air-conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, wherein In S1, the construction method of the regional air quality model is as follows: The carbon dioxide concentration is used to measure the regional air quality, and according to the mass balance equation, the regional carbon dioxide concentration balance model is obtained: C ma = drC oa + (1 - dr)C in (8) G in = G person N (9) Among them, C in represents the regional carbon dioxide volume fraction, ppm; C ma represents the mixed air carbon dioxide volume fraction, ppm; C oa represents the outdoor carbon dioxide volume fraction, ppm; G sa represents the regional supply air volume flow rate, m 3 / s; V represents the regional space volume, m 3 ; A inf represents the infiltration air rate, s -1 ; dr represents the fresh air ratio; G in represents the total rate of carbon dioxide generation by passengers in the region, m 3 / s; G person represents the rate of carbon dioxide generation by passengers, m 3 / (s·person); N represents the number of passengers in the region, person; Passenger average carbon dioxide emission rate G person The calculation formula is expressed as: Among them, A D represents the DuBois surface area, m 2 ; M is the physical activity level in Met, which is set according to the activity types of passengers in the terminal; RQ represents the respiratory quotient, which is the molecular ratio of CO2 released and O2 absorbed during respiration.

6. A double-layer optimization control method for an airport terminal air conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, characterized in that, In S2, passengers are regarded as individual movement units, and by defining the transition probability matrix under different event mechanisms for their movement modes, the displacement of passengers in the area is characterized and simulated to generate the hourly passenger occupancy situation in the sub-areas, specifically as follows: S21: Define an event mechanism based on passenger flow volume with the average value of the predicted number of passengers on the same day as the standard; S22: Obtain the number of passengers in the check-in area, security check area, and waiting area, which are three different functional areas, under the current control time domain. Based on the event mechanism defined in S21, obtain the passenger transfer probability matrix; S23: Based on the passenger flow volume under the current control time domain and the passenger transfer probability matrix in S22, calculate the passenger distribution in different air-conditioning supply areas within each functional area according to the Markov moving model, and divide the occupied area and the unoccupied area.

7. A double-layer optimal control method for an airport terminal air-conditioning system considering the spatio-temporality of passenger flow according to claim 6, characterized in that, The specific content of S21 is as follows: If the number of passengers N(t) in the functional area of the current control time domain is less than the average number of passengers on the same day then set the transfer probability p of half of the air supply areas for personnel ij to zero, and generate probability distribution random numbers for the remaining air supply areas through the Monte Carlo method. Obtain the passenger number distribution of each sub-area according to the initial state and the transfer probability matrix; otherwise, on the premise of ensuring that there are people in each air supply area, generate probability distribution random numbers through the Monte Carlo method, and obtain the passenger number distribution of different air supply areas according to the initial state and the transfer probability matrix; The specific content of S22 is as follows: Assume that the number of air-conditioning supply areas in each functional area is n, and the numbers of different air-conditioning supply areas are in the set I = {1, 2,..., n}. Then the transition probability matrix of the Markov chain is expressed as: where each element p in the matrix ij represents the probability of transitioning from state i to state j, that is, it represents the probability that a person in the area is in the air-conditioning supply area i at time t and in the air-conditioning supply area j at time t + 1; Transition probability matrix P t Satisfy: all elements are non - negative and the sum of elements in any row is 1.

8. A double-layer optimization control method for an airport terminal air conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, characterized in that, In S3, the optimization objective function includes the operating energy consumption of the air-conditioning system, the tracking performance of thermal and humidity comfort, and the tracking performance of regional air quality. The construction method is as follows: The operating energy consumption model of the air-conditioning system includes refrigeration energy consumption and fan energy consumption. The total energy consumption J1(t) is expressed as: J1(t) = P c (t) + P f (t) (12) Among them, P c (t) represents the refrigeration energy consumption, in kW; P f (t) represents the fan energy consumption, in kW; Refrigeration energy consumption P c (t) The calculation formula is as follows: h ma (t) = (1 - dr(t))h ra (t) + dr(t)h oa (t) (14) Among them, ρ a represents the air density, kg / m 3 ; G sa (t) represents the supply air volume flow rate, m 3 / s; h ma (t) represents the mixed air enthalpy value, kJ / kg; h sa (t) represents the supply air enthalpy value, kJ / kg; dr(t) represents the fresh air ratio; h ra (t) represents the return air enthalpy value, kJ / kg; h oa (t) represents the fresh air enthalpy value, kJ / kg; COP represents the refrigeration energy efficiency ratio; Fan energy consumption P f (t) The calculation formula is as follows: P f σG(t) sa σG(t) 2 (15) Among them, σ represents the penalty coefficient of the fan power function; G sa (t) represents the air supply volume flow rate, m 3 / s; The tracking target J2(t) of thermal and humidity comfort is expressed as: Among them, T in (t) represents the regional temperature, °C; T set (t) represents the regional temperature set value, °C; represents the regional relative humidity; represents the regional relative humidity set value; Relative humidity is a function of the temperature T in and the humidity W in and is calculated as follows: Among them, P w,s represents the saturated water vapor partial pressure, kPa; P w represents the water vapor partial pressure, kPa; B represents the atmospheric pressure, with a value of 101.325 kPa; the regional air quality tracking target J3 is expressed as: J3 = (C in (t) - C min ) 2 (20) Among them, C in (t) represents the regional carbon dioxide concentration, ppm; C min represents the lower limit of the carbon dioxide volume fraction, that is, the outdoor carbon dioxide volume fraction; In S3, the specific method for clarifying the optimization objectives of the area under different passenger distribution characteristics is as follows: When the area is in the occupied state, the upper-layer thermal and humidity control layer takes the multi-objective function of thermal and humidity comfort tracking and air-conditioning system operating energy consumption based on weight allocation within the prediction time domain as the optimization objective, and the lower-layer fresh air control layer takes the multi-objective function of regional air quality and air-conditioning system operating energy consumption based on weight allocation within the prediction time domain as the optimization objective, as follows: Optimization Objectives of the Thermal and Humidity Control Layer in the Occupied Zone Expressed as: Among them, i represents the current control time, h; L represents the length of the prediction time domain, h; ω1 represents the weight coefficient of the air-conditioning system operating energy consumption target of the thermal and humidity control layer; J1(t) represents the air-conditioning system operating energy consumption target value at time t; ω2 represents the weight coefficient of the thermal and humidity comfort tracking target of the thermal and humidity control layer; J2(t) represents the thermal and humidity comfort tracking target value at time t; Optimization Objectives of the Fresh Air Control Layer in the Occupied Area It is expressed as: Among them, i represents the current control time, h; L represents the length of the prediction time domain, h; ω3 represents the weight coefficient of the air-conditioning system operating energy consumption target of the fresh air control layer; J1(t) represents the air-conditioning system operating energy consumption target value at time t; ω4 represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; J3(t) represents the regional air quality tracking target value at time t; When the area is in the unoccupied state, do not consider the tracking of the temperature, humidity, and carbon dioxide concentration in the area, and only maintain them within the constraints. Only consider the optimization objective of the air-conditioning system operating energy consumption, as follows: The optimization objective of the thermal and humidity control layer in the unoccupied area is expressed as: Among them, \(i\) represents the current control moment, \(h\); \(L\) represents the prediction time domain length, \(h\); \(J_1(t)\) represents the total energy consumption of the system operation, \(kW\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\); The optimization objective of the fresh air control layer in the unoccupied area is expressed as: Among them, i represents the current control time, h; L represents the prediction time domain length, h; J1(t) represents the total energy consumption of the system operation, kW; P c (t) represents the refrigeration energy consumption, kW; P f (t) represents the fan energy consumption, kW; When calculating the air-conditioning system operating energy consumption, the fresh air ratio of the thermal and humidity control layer is a fixed value, and the remaining air supply state parameters are optimization variables. The air supply state parameters of the fresh air control layer are the optimization results of the thermal and humidity control layer, and the fresh air ratio is an optimization variable.

9. A double-layer optimization control method for a terminal air conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, characterized in that In S4, the parameter setting of the double-layer control model includes the parameters of each area, the control parameters of the model predictive control algorithm, and the initial parameters of the decision variables. The control parameters include the control time domain and the prediction time domain. Among them, the prediction time domain data includes outdoor meteorological data, the Markov transition probability matrix of each area, the number of passengers and occupancy in each area, the air-conditioning operation parameters, and the remaining heat disturbance data. The outdoor meteorological parameters include outdoor temperature, outdoor relative humidity, and solar radiation intensity. The remaining heat disturbance data includes the heat and moisture loads caused by infiltration, solar radiation, equipment, and lighting. The double-layer control model includes a heat and moisture control layer and a fresh air control layer. The constraint conditions of the heat and moisture control layer include the regional temperature change constraint, the regional relative humidity change constraint, the supply air volume flow constraint, the supply air temperature constraint, and the supply air humidity constraint. The constraint conditions of the fresh air control layer include the regional carbon dioxide concentration constraint and the fresh air ratio constraint, which are specifically as follows: The regional temperature change constraint is: Among them, represents the lower limit of the regional temperature, °C; represents the upper limit of the regional temperature, °C; T in represents the regional temperature, °C; The regional relative humidity change constraint is: Among them, represents the lower limit of the regional relative humidity; represents the upper limit of the regional relative humidity; represents the regional relative humidity; The supply air volume flow constraint is: Among them, represents the lower limit of the air supply volume flow rate, m 3 / s; represents the upper limit of the air supply volume flow rate, m 3 / s; G sa represents the air supply volume flow rate, m 3 / s; The supply air temperature constraint is: Among them, represents the lower limit of the supply air temperature, °C; represents the upper limit of the supply air temperature, °C; T sa represents the supply air temperature, °C; The supply air humidity constraint is: Among them, represents the lower limit of the supply air humidity, g / kg; represents the upper limit of the supply air humidity, g / kg; W sa represents the supply air humidity, g / kg; The regional carbon dioxide concentration constraint: Among them, represents the lower limit of the regional carbon dioxide concentration, ppm; represents the upper limit of the regional carbon dioxide concentration, ppm; C in represents the regional carbon dioxide concentration, ppm; The fresh air ratio constraint: dr min ≤dr≤dr max (31) Among them, dr min represents the lower limit of the fresh air ratio; dr max represents the upper limit of the fresh air ratio; dr represents the fresh air ratio.

10. A double-layer optimization control method for an airport terminal air conditioning system considering the spatio-temporal characteristics of passenger flow according to claim 2, characterized in that, The rolling optimization model of the double-layer control model in S5 is specifically as follows: In the occupied state, the rolling optimization model of the heat and moisture control layer is expressed as: Among them, i represents the current control moment, h; L represents the length of the prediction time domain, h; ω1 represents the weight coefficient of the heat and moisture comfort tracking target of the heat and moisture control layer; J1(t) represents the heat and moisture comfort tracking target value at time t; ω2 represents the weight coefficient of the system operation energy consumption target of the heat and moisture control layer; J2(t) represents the system operation energy consumption target value at time t. The rolling optimization model of the fresh air control layer is expressed as: Among them, i represents the current control moment, h; L represents the length of the prediction time domain, h; ω3 represents the weight coefficient of the system operation energy consumption target of the fresh air control layer; J1(t) represents the system operation energy consumption target value at time t; ω4 represents the weight coefficient of the regional air quality tracking target of the fresh air control layer; J3(t) represents the regional air quality tracking target value at time t. In the unoccupied state, the rolling optimization model of the heat and moisture control layer is expressed as: where \(i\) represents the current control time, \(h\); \(L\) represents the prediction horizon length, \(h\); \(P c (t)\) represents the refrigeration energy consumption, \(kW\); \(P f (t)\) represents the fan energy consumption, \(kW\); The rolling optimization model of the fresh air control layer is expressed as: where \(i\) represents the current control time, \(h\); \(L\) represents the prediction time domain length, \(h\); \(P\) c (t) represents the refrigeration energy consumption, \(kW\); \(P\) f (t) represents the fan energy consumption, \(kW\).

Images