Dynamic energy management method and system for central air conditioning system based on large model

Abstract

The invention relates to the technical field of energy management, in particular to a dynamic energy management method and system for a central air-conditioning system based on a large model, and the method specifically comprises the steps: outputting a short-term load demand prediction curve in a rolling manner through a time sequence depth prediction model; carrying out inversion calculation on equivalent heat capacity and equivalent heat resistance of the building material through a physical information neural network, and quantifying thermal inertia parameters of the building material in real time; evaluating a phase change delay time coefficient of the building material; obtaining a photovoltaic power generation power prediction curve of a power grid system and a power grid time-of-use electricity price signal, and determining an optimal energy charging and discharging time window of the building material; and adopting a multi-objective decision algorithm to decide and generate an optimal operation strategy of the central air-conditioning system. The problem that in the prior art, an energy management system matched with the thermal dynamic characteristics of a building is lacked is solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of energy management, and is a central air conditioning system dynamic energy management method and system based on a large model. BACKGROUND

[0002] The current energy management strategy of the central air conditioning system has serious adaptability problems when facing modern buildings using new high-performance energy-saving building materials, such as phase change materials PCM and high-heat-capacity concrete. Traditional building management systems (BMS) mostly rely on simple feedback control mechanisms (such as PID control) or static load prediction models based on historical meteorological data and operation records. The time scale of the control response is minute-level, while the thermal dynamic process of high-thermal-inertia building materials shows a slow response of hours or even days. This fundamental time scale mismatch leads to significant control lag and over-regulation of the system, which is specifically manifested in that: at the initial stage of refrigeration demand, the air conditioning output is difficult to quickly affect the indoor environment, and after the refrigeration equipment is turned off, the heat or cold stored in the building materials is still released, causing temperature fluctuations and energy waste. In addition, conventional prediction models do not consider the thermal inertia parameters of building structures and their time-varying characteristics, which cannot accurately quantify the thermal energy storage effect of building materials and cannot effectively coordinate the operation of central air conditioning and intermittent renewable energy (such as photovoltaic), resulting in the inability to achieve effective peak shaving and valley filling under the mechanism of time-of-use electricity price, and instead, the improper control may cause adverse effects, such as forced high-load operation during the peak period of night-time electricity price to offset the release of building material heat storage, causing double losses in economy and energy efficiency. SUMMARY

[0003] The application aims to solve the problem in the prior art that there is a lack of energy management systems matched with the thermal dynamic characteristics of buildings, and provides a central air conditioning system dynamic energy management method and system based on a large model.

[0004] To achieve the above purpose, the technical scheme of the central air conditioning system dynamic energy management method based on a large model comprises the following steps: S1: fuse real-time indoor and outdoor environment data and central air conditioning equipment state data, and roll out a short-term load demand prediction curve through a time series deep prediction model; S2: collect temperature field data of building materials, calculate the equivalent heat capacity and equivalent thermal resistance of the building materials through a physical information neural network, and quantize the thermal inertia parameters of the building materials in real time; S3: based on the real-time quantized thermal inertia parameters of the building materials in S2, predict the temperature field change trend of the building structure in the next 3 hours, and based on the predicted temperature field change trend, evaluate the phase change delay time coefficient of the building materials; S4: Obtain the photovoltaic power prediction curve of the power grid system and the time-of-use electricity price signal of the power grid, and determine the optimal energy charging and discharging time window of the building materials; S5: According to S1-S4, a multi-objective decision algorithm is used to generate the optimal operation strategy of the central air conditioning system.

[0005] Preferably, in S1, the short-term refined load demand prediction curve is a refined load demand prediction curve for several hours in the future; The acquisition strategy of the short-term refined load demand prediction curve comprises: S11: According to real-time indoor and outdoor environmental data and equipment state data of the central air conditioner, a feature weighting fusion is performed using an attention mechanism to dynamically correct the short-term load prediction result; S12: A multi-scale convolutional neural network is used to extract periodic patterns, trend components and abnormal fluctuation features in the historical load sequence, and the future 6-hour load prediction value is updated in 1-hour intervals to obtain the short-term load demand prediction curve .

[0006] Preferably, S2 comprises the following steps: S21: Collecting temperature sensor data of the east, south, west and north facades and the roof inside the building, as well as corresponding external surface solar radiation intensity data, to form a multi-dimensional thermal state sequence, and performing data standardization processing according to the differences in solar radiation of different facades; S22: Inputting the above multi-dimensional sequence into a physical information neural network, wherein the physical information neural network takes the unsteady heat conduction equation as a physical constraint to construct a loss function , and optimizing the phase change material characteristics of the building materials through the loss function ; S23: Inverting the equivalent thermal inertia parameters of the building materials by minimizing the loss function , wherein the equivalent thermal inertia parameters comprise an equivalent heat capacity and an equivalent thermal resistance ; S24: Extracting the equivalent thermal inertia parameters output by step S23, and using the equivalent heat capacity divided by the equivalent thermal resistance to obtain the real-time thermal inertia index of the building materials; S25: Real-time monitoring the phase change state of the building materials, and calculating the phase change degree of the building materials ; Based on the phase change degree of the building materials , calculating the latent heat absorption and release rate .

[0007] Preferably, S3 comprises: S31: Based on equivalent thermal inertia parameters, the building structure is simplified into a thermal network model consisting of thermal resistance and heat capacity nodes. According to the law of conservation of energy, the discrete-time state-space differential equation of the thermal network model is established and solved using the Euler algorithm to predict the temperature field trend of the building structure in the next three hours. S32: Evaluate the phase change delay time coefficient of building materials based on the predicted temperature field change trend .

[0008] Preferably, S4 includes the following steps: S41: Extracting the real-time thermal inertia index of the building material output in step S25 , and according to the preset thermal inertia index threshold Conduct thermal response capability level analysis and verification of building materials. When the real-time thermal inertia index of building materials is Greater than or equal to the thermal inertia index threshold When it is determined that the building has strong thermal inertia, predictive energy storage control is adopted and step S42 is continued; When the real-time thermal inertia index of building materials Less than the thermal inertia index threshold When , it is judged that the building thermal inertia is weak and the standard predictive control strategy is continued; S42: Preset threshold value for effective utilization rate of latent heat of phase change materials in building materials and phase change delay time coefficient threshold ; S43: intercepting the load forecast curve for the next three hours from the load demand forecast curve in step S1, extracting the peak load period and the valley load period therein to form a peak load set and a low-peak load set respectively, and identifying the photovoltaic power generation peak period and the grid electricity price peak period; S44: Filter peak load sets with latent heat utilization rates lower than or a delay greater than The number of time periods filtered out is ; S45: According to the thermal inertia parameters of building materials and the number of time periods obtained by screening, Calculate the optimal charging and discharging time window for building materials, specifically: ; in, They are the optimal charging time and the optimal releasing time of building materials; are the starting moments of the grid low price and grid high price, respectively; They are the difference between the building’s indoor set temperature and the maximum temperature allowed for the building structure to operate; is the temperature difference between the inside and outside of the building at the current time; is the thermal inertia time constant of the building material, which is the equivalent heat capacity output in step S23 and the equivalent thermal resistance ; is a time delay parameter determined according to the material properties of the building material.

[0009] Preferably, S51: according to S41-S45, the control parameters to be updated of the central air conditioning system are calculated, and the control parameter adjustment strategy is as follows: ; wherein k is the number of iterations of the control parameter optimization; is the comprehensive energy efficiency ratio of the central air conditioning system in the kth iteration; is the new comprehensive energy efficiency ratio target value of the central air conditioning system in the next iteration; Z is the total number of time periods; is the preset threshold value of the number of time periods; is the weight coefficient of the influence of the thermal characteristics of the building material; S52: extract and store the updated comprehensive energy efficiency ratio target value of the optimized central air conditioning system in the cloud database, and simultaneously issue it to the local controllers of the refrigeration host, variable frequency water pump and cooling tower of the central air conditioning through the industrial communication protocol.

[0010] S53: extract the power generation data of the photovoltaic power generation peak period to form a photovoltaic output set, and simultaneously extract the air conditioning refrigeration load data of the same period to form an air conditioning load set, and distinguish between the basic load and the adjustable load; S54: adopt an energy management strategy based on the thermal inertia of the building material, which specifically includes: evaluating the target cold storage capacity of the building material , and monitoring the cumulative cold storage capacity of the building material in real time; When the cumulative cold storage capacity is lower than the target cold storage capacity , increase the operating power of the central air conditioning system during the photovoltaic power generation peak period, and store the excess cold in the high thermal inertia building material; When the cumulative cold storage capacity is equal to or higher than the target cold storage capacity , it means that the cold storage of the building material is completed, and the operating power of the central air conditioning system is adjusted to the normal working level; wherein the calculation strategy of the target cold storage capacity is: ; wherein, is the photovoltaic power at time t; is the basic load power at time t; is the latent heat absorbed or released during the phase change of the building material; S55: Evaluate from the mismatch degree of photovoltaic output, air conditioning load, and heat capacity in three different dimensions respectively to obtain photovoltaic evaluation items, air conditioning load evaluation items, and heat capacity matching items, and synchronously evaluate photovoltaic-air conditioning load matching degrees from the three dimensions .

[0011] S56: Extract photovoltaic-air conditioning load matching degrees , perform switching judgment of the central air conditioning working mode, including: a preset matching degree threshold , and perform collaborative level analysis according to the matching degrees, when the photovoltaic-air conditioning load matching degree is greater than or equal to , trigger a collaborative control execution mode; when the photovoltaic-air conditioning load matching degree is less than , do not trigger the collaborative control execution unit, and start a backup power grid power supply strategy; the collaborative control execution mode includes: increasing the air conditioning system operation power during a photovoltaic output peak period, storing excess cold energy in a building structure, and releasing the stored cold energy during a peak electricity price period.

[0012] In addition, the dynamic energy management system of the central air conditioning system based on a large model includes the following modules: a load prediction module, a thermal inertia quantification module, a phase change quantification module, a dynamic energy storage control module, and an air conditioning operation optimization module; the load prediction module is used for fusing real-time indoor and outdoor environment data and device state data of the central air conditioning, and rolling outputting a short-term load demand prediction curve through a time series deep prediction model; the thermal inertia quantification module is used for collecting temperature field data of building materials, inversely calculating the equivalent heat capacity and equivalent thermal resistance of the building materials through a physical information neural network, and quantifying the thermal inertia parameters of the building materials in real time; the phase change quantification module is used for predicting the temperature field change trend of the building structure in the next 3 hours, and evaluating the phase change delay time coefficient of the building materials based on the predicted temperature field change trend; the dynamic energy storage control module is used for obtaining a photovoltaic power generation power prediction curve of a power grid system and a power grid time-of-use electricity price signal, and determining the best energy charging and discharging time window of the building materials; the air conditioning operation optimization module adopts a multi-objective decision algorithm to decide and generate an optimal operation strategy of the central air conditioning system.

[0013] Compared with the prior art, the technical effects of the present application are as follows: 1. The present application inverses the equivalent heat capacity and equivalent thermal resistance of the building structure online through physical information neural network, and evaluates the phase change degree of building materials in real time, fundamentally solving the serious control lag and excessive response problem caused by the inability to quantify the thermal dynamic characteristics of building materials in the existing control method, enabling the system to accurately predict the future trend of building structure temperature field changes for several hours, and calculate the optimal energy charging and discharging time window in advance based on the phase change delay time constant, thereby completely avoiding room temperature fluctuations and energy waste, and greatly improving indoor thermal comfort and system operation stability. 2. The present application performs multi-objective collaborative optimization on photovoltaic power generation prediction, grid time-of-use price signals and building energy storage characteristics, actively increases air conditioner operation power to store cold energy during the peak period of photovoltaic output, and releases the stored cold energy to reduce grid power purchase demand during the peak period of electricity price, which not only greatly improves the self-generation and self-use rate of photovoltaic energy, but also achieves significant peak shaving and valley filling effect, effectively reduces system operation cost, and at the same time introduces photovoltaic-air conditioner load matching degree index and threshold judgment mechanism to intelligently switch between collaborative control and grid power supply mode, ensuring the reliability and economy of energy dispatching strategy. 3. The present application dynamically coordinates the operation state of air conditioner host, water pump and other equipment through the control parameter self-adaptive adjustment mechanism with system comprehensive energy efficiency ratio as the core, so that they are always in the high-efficiency working interval, and more comprehensively realizes the multi-objective balance of energy consumption economy, indoor comfort and demand response income. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them: Figure 1 The flowchart of a central air conditioning system dynamic energy management method based on a large model of the present application; Figure 2 The structural diagram of a central air conditioning system dynamic energy management system based on a large model of the present application. DETAILED DESCRIPTION

[0015] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0016] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0017] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0018] Example 1: like Figure 1 As shown, a dynamic energy management method for a central air-conditioning system based on a large model according to an embodiment of the present invention is as follows: Figure 1 As shown, the specific steps are as follows: S1: Integrates real-time indoor and outdoor environmental data with central air conditioning equipment status data, and outputs a short-term load demand forecast curve using a time series deep prediction model. In S1, the short-term refined load demand forecast curve is a refined load demand forecast curve for several hours in the future; The strategy for obtaining the short-term refined load demand forecast curve includes: S11: Based on real-time indoor and outdoor environmental data and central air conditioning equipment status data, an attention mechanism is used to perform weighted feature fusion and dynamically correct short-term load forecast results; S12: A multi-scale convolutional neural network is used to extract the periodic patterns, trend components, and abnormal fluctuation characteristics from the historical load series, and the load forecast value for the next 6 hours is updated at 1-hour intervals to obtain a short-term load demand forecast curve. .

[0019] Exemplarily, in this embodiment, the temporal depth prediction model adopts a hybrid architecture combining a temporal convolutional network and a Transformer attention mechanism; S2: Collect temperature field data of building materials, calculate the equivalent heat capacity and equivalent thermal resistance of building materials through physical information neural network inversion, and quantify the thermal inertia parameters of building materials in real time; S2 includes the following steps: S21: Collect temperature sensor data from the east, south, west, and north facades and the roof, as well as the corresponding external surface solar radiation intensity data, to form a multi-dimensional thermal state sequence. Perform data standardization based on the differences in solar radiation received by different facades. S22: input the multi-dimensional sequence into a physical information neural network, the physical information neural network taking a non-steady-state heat conduction equation as a physical constraint to construct a loss function , the loss function is optimized according to the phase change material characteristics of the building materials; Exemplarily, in the embodiment, the construction of the loss function includes: simultaneously constructing a first loss term , a second loss term and a third loss term , and performing normalization and weighted summation processing on the first loss term , the second loss term and the third loss term through preset first, second and third weight coefficients, to finally obtain the loss function ; The first loss term is: ; Wherein, N is the number of temperature measurement points; represents the temperature predicted by the physical information neural network at the i-th measurement point; represents the temperature actually measured at the i-th point; It should be noted that the first loss term is a data fitting term, which aims to quantify the prediction error of the physical information neural network; The second loss term is: ; Wherein, M is the number of configuration points randomly selected in the calculation domain, used to verify whether the physical information neural network satisfies the heat conduction rule; is the partial derivative of temperature with respect to time; is the thermal diffusivity of the building materials; is the Laplacian of temperature, which represents the spatial temperature distribution non-uniformity; It should be noted that the second loss term is a physical constraint term, which is constructed based on the partial differential equation in the embodiment, representing the instantaneous change rate of the temperature of a point inside an object, which is proportional to the temperature distribution around the point; The third loss term is: ; It should be noted that the third loss term is a phase change characteristic term; wherein K is the number of monitoring points for monitoring the phase change degree of the building materials; is the predicted value of the phase change degree of the building materials, is the actually measured phase change degree of the building materials; S23: By minimizing the loss function The equivalent thermal inertia parameters of building materials are obtained by inversion, and the equivalent thermal inertia parameters include: equivalent heat capacity and equivalent thermal resistance ; S24: Extract the equivalent thermal inertia parameters output from step S23 and use the equivalent heat capacity Divide by the equivalent thermal resistance Get the real-time thermal inertia index of building materials ; It should be noted that the thermal inertia index of building materials Used to quantify the heat storage and release capabilities of new building materials; S25: Real-time monitoring of the phase change state of building materials and calculation of the phase change degree of building materials ; Based on the phase change degree of building materials Calculate the latent heat absorption and release rate ; For example, in this embodiment, a phase change degree of a building material is provided. The calculation strategy is as follows: ; in, is the temperature measurement value of the i-th temperature measurement point, is the phase transition temperature of building materials, k is the phase transition slope coefficient; It should be noted that, considering that the phase change process of building materials is not completed instantaneously at a precise temperature point, but occurs gradually within a temperature range, in this embodiment, the phase change process of building materials is described by a Sigmoid function. It should also be noted that, considering that the temperature distribution of phase change building materials in building components is not completely uniform, averaging is used to quantify the spatial average of the phase change degree of building materials.

[0020] For example, based on the law of conservation of energy, in this embodiment, a method based on the phase change degree of building materials is also provided. Calculate latent heat absorption and release rate Implementation examples are as follows: ; in, is the density of building materials, V is the total volume of building materials, and L is the latent heat value; It should be noted that It indicates the heat power absorbed or released by building materials due to phase change per unit time; S3: based on the real-time quantification of S2, the thermal inertia parameter of the building material, predict the temperature field change trend of the building structure in the next 3 hours, and based on the predicted temperature field change trend, evaluate the phase change delay time coefficient of the building material; S3 includes: S31: based on the equivalent thermal inertia parameter, simplify the building structure into a thermal network model composed of a thermal resistance and a thermal capacity node, establish the differential equation of the discrete time state space of the thermal network model according to the law of conservation of energy, and solve it by Euler algorithm to predict the temperature field change trend of the building structure in the next 3 hours; Exemplarily, in the embodiment, a prediction implementation example of the temperature field change trend of the building structure in the next 3 hours is provided, which includes: In this embodiment, the building is regarded as a uniform whole point, that is, a first-order RC model, and the equivalent thermal capacity and the equivalent thermal resistance are calculated. As the resistance and capacitance of the first-order RC model, the outdoor temperature and the indoor temperature of the building are regarded as the voltage source and the voltage across the capacitor of the first-order RC model. According to the law of conservation of energy, the differential equation of the discrete time state space is established, which is specifically: For the above differential equation, in this embodiment, it needs to be explained that the left side of the differential equation is intended to represent the instantaneous power of the energy increase in the building; and the right side of the differential equation is intended to represent all external sources causing the change of the energy in the building, in which is the heat intrusion power due to the outdoor temperature, is the refrigeration power of the central air conditioning when refrigerating, which represents the process of removing heat from the building, so subtraction is used to quantify the process of removing heat by the central air conditioning; The current indoor and outdoor temperature and the internal temperature monitoring value of the building material are taken as the initial state, the future meteorological prediction data and the running set value are taken as the input, and the above differential equation is recursively solved by Euler algorithm to predict the temperature field change trend of the building structure in the next 3 hours.

[0021] S32: based on the predicted temperature field change trend, evaluate the phase change delay time coefficient of the building material .

[0022] Exemplarily, in the embodiment, a prediction implementation example of the temperature field change trend of the building structure in the next 3 hours is provided, which includes: ​evaluation strategy, specifically: ; wherein, is the actual temperature change curve, is the temperature curve of the temperature field change trend predicted in step S31, is the lowest temperature measurement value in the observation period; is the observation period; It should be noted that in the present embodiment, the phase change delay time coefficient of the building material is evaluated is intended to quantify the thermal response hysteresis of the building material.

[0023] S4: Obtain the photovoltaic power generation prediction curve of the power grid system and the time-of-use electricity price signal, and determine the optimal energy charging and discharging time window of the building material; S4 includes the following steps: S41: Extract the real-time thermal inertia index of the building material output in step S25 , and perform thermal response capability grade analysis and verification of the building material according to a preset thermal inertia index threshold When the real-time thermal inertia index of the building material is greater than or equal to the thermal inertia index threshold , it is judged that the building thermal inertia is strong, and the predictive energy storage control is adopted, and step S42 is continued to be executed; When the real-time thermal inertia index of the building material is less than the thermal inertia index threshold , it is judged that the building thermal inertia is weak, and the standard prediction control strategy is continued to be adopted.

[0024] S42: Pre-set the latent heat effective utilization rate threshold and the phase change delay time coefficient threshold of the phase change material in the building material; Exemplarily, in the present embodiment, a preset strategy for the latent heat effective utilization rate threshold is provided, specifically: ; wherein, is the maximum theoretical latent heat absorption or release rate determined according to the material properties of the building material; It should be noted that is used to quantify the percentage of the latent heat capacity of the building material that is effectively utilized in actual operation; represents the cumulative latent heat absorbed or released by the building material in the actual observation period; represents the maximum theoretical latent heat that can be absorbed or released by the building material in the same observation period.

[0025] S43: intercepting the load forecast curve for the next three hours from the load demand forecast curve in step S1, extracting the peak load period and the valley load period therein to form a peak load set and a low-peak load set respectively, and identifying the photovoltaic power generation peak period and the grid electricity price peak period; S44: Filter peak load sets with latent heat utilization rates lower than or a delay greater than The number of time periods filtered out is ; S45: According to the thermal inertia parameters of building materials and the number of time periods obtained by screening, Calculate the optimal charging and discharging time window for building materials, specifically: ; in, They are the optimal charging time and the optimal releasing time of building materials; are the starting moments of the grid low price and grid high price, respectively; They are the difference between the building’s indoor set temperature and the maximum temperature allowed for the building structure to operate; is the temperature difference between the inside and outside of the building at the current moment; is the thermal inertia time constant of building materials, which is the equivalent heat capacity output in step S23. and equivalent thermal resistance The product of It is a time delay parameter determined according to the material properties of building materials.

[0026] It should be noted that building materials have inherent thermal inertia time constants. , there is also an additional response delay during phase change, i.e. ; It should be noted that in this embodiment, the starting times of low and high grid prices are used as the basis. Specifically, considering the strong thermal inertia of building materials and their response to temperature, the time term generated by subtracting the response of building materials from the starting time of low grid prices is used. This is to extract the energy absorption state of building materials before the starting time of low grid prices, so that they can reach the expected state at the target time (i.e., when low-price electricity starts); For the Aims to quantify the thermal inertia time constant of building materials required to change from the current state to the target state; S5: Based on S1-S4, a multi-objective decision-making algorithm is used to generate the optimal operation strategy of the central air-conditioning system.

[0027] S5 includes: S51: According to S41-S45, the control parameters of the central air conditioning system to be updated are calculated, and the control parameter adjustment strategy is as follows: ; Wherein, k is the number of iterations of the control parameter optimization; is the comprehensive energy efficiency ratio of the central air conditioning system in the kth iteration; is the new comprehensive energy efficiency ratio target value of the central air conditioning system in the next iteration; Z is the total number of time periods; is the preset threshold of the number of time periods; is the weight coefficient of the influence of building materials thermal characteristics; In this embodiment, it should be noted that the is intended to represent how many time periods the performance of building materials in storing or releasing energy is not up to standard; When , it means that these time periods cannot independently meet the future load demand in the building by building materials, so the operating energy efficiency target COP of the central air conditioning system needs to be improved to compensate for the insufficient energy storage of building materials; And when , it means that these time periods can independently meet the future load demand in the building by building materials, so the operating energy efficiency target COP of the central air conditioning system needs to be reduced to save energy.

[0028] S52: Extract and store the updated comprehensive energy efficiency ratio target value of the optimized central air conditioning system in the cloud database, and issue it to the local controllers of the refrigeration main machine, variable frequency water pump and cooling tower of the central air conditioning through the industrial communication protocol.

[0029] Exemplarily, in this embodiment, after receiving the updated comprehensive energy efficiency ratio target value, the local controllers of the refrigeration main machine, variable frequency water pump and cooling tower of the central air conditioning call their internal preset optimization algorithm to autonomously solve and adjust their local execution equipment to the optimal working point. Exemplarily, the optimization algorithm is fuzzy PID; In this embodiment, it should also be noted that, in order to avoid frequent start and stop of the equipment, a smooth transition algorithm is used when adjusting the frequency of the refrigeration main machine.

[0030] S5 also includes: S53: Extract the power generation data of the photovoltaic power generation peak period to form a photovoltaic output set, and simultaneously extract the air conditioning refrigeration load data of the same period to form an air conditioning load set, and distinguish the basic load and the adjustable load; S54: Adopt an energy management strategy based on the thermal inertia of building materials, which specifically includes: evaluating the target cold storage capacity of building materials , and monitoring the cumulative cold storage capacity of building materials in real time; When the cumulative cold storage amount is lower than the target cold storage amount , the operation power of the central air conditioning system is increased during the peak period of photovoltaic power generation, and the excess cold is stored in the high-thermal-inertia building materials; When the cumulative cold storage amount is equal to or higher than the target cold storage amount , it indicates that the cold storage of the building materials is completed, and the operation power of the central air conditioning system is adjusted to the normal working level; The target cold storage amount is calculated as follows: ; Wherein, is the photovoltaic power generation power at time t; is the basic load power at time t; is the latent heat absorbed or released during the phase change of the building materials; In this embodiment, The acquisition strategy is as follows: the latent heat absorption and release rate output by step S25 is extracted, and is integrated in the period from the beginning of the cold storage of the building materials to the current time to obtain ; S55: Evaluate the mismatch degree from three different dimensions of photovoltaic output, air conditioning load, and heat capacity respectively, obtain the photovoltaic evaluation item, the air conditioning load evaluation item, and the heat capacity matching item, and evaluate the photovoltaic-air conditioning load matching degree from the three dimensions synchronously .

[0031] Exemplarily, in this embodiment, the calculation strategy of the photovoltaic-air conditioning load matching degree is as follows: evaluate the mismatch degree from three different dimensions of photovoltaic output, air conditioning load, and heat capacity respectively, obtain the photovoltaic evaluation item, the air conditioning load evaluation item, and the heat capacity matching item, and then perform weighted summation on the three evaluation items to obtain the photovoltaic-air conditioning load matching degree ; Wherein, are the photovoltaic evaluation item, the air conditioning load evaluation item, and the heat capacity matching item respectively; In this embodiment, ; ; Wherein, is the photovoltaic power generation power at time t; is the air conditioning refrigeration demand power at time t predicted by step S1; are the maximum and minimum values of the photovoltaic power generation power respectively; are the maximum and minimum values of the air conditioning refrigeration power respectively; In this embodiment, ; wherein, is the available heat capacity of the building structure at time t, which represents how much heat the building structure can still absorb or release at time t; is the required heat capacity at time t, which represents how much heat the building structure still needs to absorb or release from time t to meet the comfort requirements inside the building, i.e. the central air conditioner can meet the predicted load demand; is the maximum heat capacity that the building structure can store.

[0032] It should be noted that in the present embodiment, the acquisition strategy is to extract the measured average temperature of the current building structure, first subtract the upper limit of the operating temperature range allowed by the building structure from the measured average temperature, and then multiply the equivalent heat capacity by the difference obtained above to obtain ; It should be noted that in the present embodiment, the acquisition strategy is to obtain the target set temperature of the current central air conditioner, and simultaneously extract the temperature field change trend of the building structure predicted in step S31 for the next 3 hours, calculate the difference between the target set temperature and the predicted temperature in the temperature field change trend, and multiply the equivalent heat capacity by the difference obtained above to obtain ; It should be noted that in the present embodiment, the value of is the product of the equivalent heat capacity output in step S23 and the operating temperature range allowed by the building structure.

[0033] S56: Extract the photovoltaic-air conditioner load matching degree to make a switching judgment of the working mode of the central air conditioner, including: a preset matching degree threshold , and perform a collaborative level analysis according to the matching degree, when the photovoltaic-air conditioner load matching degree is greater than or equal to , trigger the collaborative control execution mode; when the photovoltaic-air conditioner load matching degree is less than , do not trigger the collaborative control execution unit, and start the backup power grid power supply strategy; The collaborative control execution mode includes: increasing the operating power of the air conditioning system during the peak period of photovoltaic output, storing the excess cold in the building structure, and releasing the stored cold during the peak period of electricity price to reduce the electricity purchase cost of the power grid; It should be further noted in the present embodiment that the charging and discharging process of the building materials adopts a fuzzy control algorithm to ensure that the indoor temperature fluctuation is within the comfort range.

[0034] Embodiment Two As shown in the figure, a large model-based central air conditioning system dynamic energy management system of an embodiment of the application includes the following modules: Figure 2 Figure 2 A load prediction module, a thermal inertia quantification module, a phase change quantification module, a dynamic energy storage control module, and an air conditioning operation optimization module. The load prediction module is configured to fuse real-time indoor and outdoor environment data and central air conditioning equipment state data, and output a short-term load demand prediction curve through a time series deep prediction model. The thermal inertia quantification module is configured to collect temperature field data of building materials, calculate the equivalent heat capacity and equivalent thermal resistance of the building materials through a physical information neural network, and quantize the thermal inertia parameters of the building materials in real time. The phase change quantification module is configured to predict the temperature field change trend of the building structure in the next 3 hours, and evaluate the phase change delay time coefficient of the building materials based on the predicted temperature field change trend. The dynamic energy storage control module is configured to obtain a photovoltaic power generation prediction curve of a power grid system and a power grid time-of-use price signal, and determine the optimal energy charging and discharging time window of the building materials. The air conditioning operation optimization module adopts a multi-objective decision algorithm to determine the optimal operation strategy of the central air conditioning system.

[0035] Embodiment Three The embodiment provides an electronic device, including a processor and a memory, wherein the memory stores a computer program that can be called by the processor. The processor executes the above-mentioned large model-based central air conditioning system dynamic energy management method by calling the computer program stored in the memory.

[0036] The electronic device can have great differences due to different configurations or performances, and can include one or more processors (CPUs) and one or more memories, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the large model-based central air conditioning system dynamic energy management method provided by the above-mentioned method embodiment. The electronic device can also include other components for realizing device functions, for example, the electronic device can also have a wired or wireless network interface and an input and output interface, etc., to input and output data. This embodiment will not be described here.

[0037] Embodiment Four The embodiment provides a computer readable storage medium, which stores an erasable computer program.​​ When the computer program runs on the computer device, the computer device is caused to perform the above-mentioned large model-based central air conditioning system dynamic energy management method.

[0038] For example, the computer readable storage medium can be a Read-Only Memory (ROM), a Random Access Memory (RAM), a Compact Disc Read-Only Memory (CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0039] It should be understood that the size of the sequence number of each process described above in various embodiments of the present application does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0040] It should be understood that determining B according to A does not mean that B is determined only according to A, but B can also be determined according to A and / or other information.

[0041] The above-mentioned embodiments can be realized by software, hardware, firmware or any combination thereof, in whole or in part. When realized by software, the above-mentioned embodiments can be realized in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the flow or function according to the embodiments of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transferred from one computer readable storage medium to another, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through a wired network or / and a wireless network. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (e.g. floppy disk, hard disk, magnetic tape), an optical medium (e.g. DVD), or a semiconductor medium. The semiconductor medium can be a solid state disk.

[0042] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software mode depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0043] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0044] In several embodiments provided by the present application, it should be understood that the disclosed system, device and method can be realized by other ways. For example, the device embodiments described above are only schematic, and the division of units is only one, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0045] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0046] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0047] In the description of the specification, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in combination with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

Claims

1. A dynamic energy management method for a central air-conditioning system based on a large model, characterized in that: The method comprises: S1: Integrates real-time indoor and outdoor environmental data with central air conditioning equipment status data, and outputs a short-term load demand forecast curve using a time series deep prediction model. S2: Collect temperature field data of building materials, calculate the equivalent heat capacity and equivalent thermal resistance of building materials through physical information neural network inversion, and quantify the thermal inertia parameters of building materials in real time; S3: Based on the real-time quantification of the thermal inertia parameters of building materials in S2, the temperature field change trend of the building structure in the next 3 hours is predicted. At the same time, based on the predicted temperature field change trend, the phase change delay time coefficient of the building materials is evaluated; S4: Obtain the photovoltaic power generation power prediction curve of the power grid system and the power grid time-of-use electricity price signal to determine the optimal charging and discharging time window for building materials; S5: Based on S1-S4, a multi-objective decision-making algorithm is used to generate the optimal operation strategy of the central air-conditioning system.

2. A dynamic energy management method for a central air-conditioning system based on a large model according to claim 1, characterized in that: In S1, the short-term refined load demand forecast curve is a refined load demand forecast curve for several hours in the future; The strategy for obtaining the short-term refined load demand forecast curve includes: S11: Based on real-time indoor and outdoor environmental data and central air conditioning equipment status data, an attention mechanism is used to perform weighted feature fusion and dynamically correct short-term load forecast results; S12: A multi-scale convolutional neural network is used to extract the periodic patterns, trend components, and abnormal fluctuation characteristics from the historical load series, and the load forecast value for the next 6 hours is updated at 1-hour intervals to obtain a short-term load demand forecast curve. .

3. A dynamic energy management method for a central air-conditioning system based on a large model according to claim 2, characterized in that: S2 includes the following steps: S21: Collect temperature sensor data from the east, south, west, and north facades and the roof, as well as the corresponding external surface solar radiation intensity data, to form a multi-dimensional thermal state sequence. Perform data standardization based on the differences in solar radiation received by different facades. S22: Input the multidimensional sequence into the physical information neural network, which uses the non-steady-state heat conduction equation as a physical constraint to construct a loss function , through the loss function Optimize the phase change material properties of building materials; S23: By minimizing the loss function The equivalent thermal inertia parameters of building materials are obtained by inversion, and the equivalent thermal inertia parameters include: equivalent heat capacity and equivalent thermal resistance ; S24: Extract the equivalent thermal inertia parameters output from step S23 and use the equivalent heat capacity Divide by the equivalent thermal resistance Get the real-time thermal inertia index of building materials ; S25: Real-time monitoring of the phase change state of building materials and calculation of the phase change degree of building materials ; Based on the phase change degree of building materials Calculate latent heat absorption and release rate .

4. A dynamic energy management method for a central air-conditioning system based on a large model according to claim 3, characterized in that S3 include: S31: Based on equivalent thermal inertia parameters, the building structure is simplified into a thermal network model consisting of thermal resistance and heat capacity nodes. According to the law of conservation of energy, the discrete-time state-space differential equation of the thermal network model is established and solved using the Euler algorithm to predict the temperature field trend of the building structure in the next three hours. S32: Evaluate the phase change delay time coefficient of building materials based on the predicted temperature field change trend .

5. A large model-based dynamic energy management method for central air-conditioning systems according to claim 4, characterized in that: S4 includes the following steps: S41: Extracting the real-time thermal inertia index of the building material output in step S25 , and according to the preset thermal inertia index threshold Conduct thermal response capability level analysis and verification of building materials. When the real-time thermal inertia index of building materials is Greater than or equal to the thermal inertia index threshold When it is determined that the building has strong thermal inertia, predictive energy storage control is adopted and step S42 is continued; When the real-time thermal inertia index of building materials Less than the thermal inertia index threshold When , it is judged that the building thermal inertia is weak and the standard predictive control strategy is continued; S42: Preset threshold value for effective utilization rate of latent heat of phase change materials in building materials and phase change delay time coefficient threshold ; S43: intercepting the load forecast curve for the next three hours from the load demand forecast curve in step S1, extracting the peak load period and the valley load period therein to form a peak load set and a low-peak load set respectively, and identifying the photovoltaic power generation peak period and the grid electricity price peak period; S44: Filter peak load sets with latent heat utilization rates lower than or a delay greater than The number of time periods filtered out is ; S45: According to the thermal inertia parameters of building materials and the number of time periods obtained by screening, Calculate the optimal charging and discharging time window for building materials, specifically: ； in, They are the optimal charging time and the optimal releasing time of building materials; are the starting moments of the grid low price and grid high price, respectively; They are the difference between the building’s indoor set temperature and the maximum temperature allowed for the building structure to operate; is the temperature difference between the inside and outside of the building at the current moment; is the thermal inertia time constant of building materials, which is the equivalent heat capacity output in step S23. and equivalent thermal resistance The product of It is a time delay parameter determined according to the material properties of building materials.

6. A dynamic energy management method for a central air-conditioning system based on a large model according to claim 5, characterized in that S5 include: S51: According to S41-S45, the control parameters to be updated of the central air-conditioning system are calculated. The control parameter adjustment strategy is as follows: ； Wherein, k is the number of iterative optimization of control parameters; is the comprehensive energy efficiency ratio of the central air-conditioning system at the kth iteration; The new comprehensive energy efficiency ratio target value of the central air-conditioning system in the next iteration; Z is the total number of time periods; is the preset time period threshold; is the weight coefficient of the influence of thermal characteristics of building materials; S52: Extract and store the updated comprehensive energy efficiency ratio target value of the optimized central air-conditioning system to the cloud database, and at the same time send it to the local controllers of the central air-conditioning refrigeration host, variable frequency water pump and cooling tower through the industrial communication protocol.

7. A large model-based dynamic energy management method for central air-conditioning systems according to claim 6, characterized in that: The S5 also includes: S53: Extracting the power generation data of the photovoltaic power generation peak period to form a photovoltaic output set, and simultaneously extracting the air conditioning cooling load data of the same period to form an air conditioning load set, and distinguishing between the base load and the adjustable load; S54: Adopt energy management strategies based on the thermal inertia of building materials, including: evaluating the target cooling capacity of building materials , real-time monitoring of the cumulative cold storage capacity of building materials; When the cumulative cooling capacity is lower than the target cooling capacity When the photovoltaic power generation peak period, the central air-conditioning system operating power is increased, and the excess cooling capacity is stored in the high thermal inert building materials; When the cumulative cooling capacity is equal to or higher than the target cooling capacity When the cooling capacity of the building materials is completed, the operating power of the central air-conditioning system is adjusted to the normal working level; Wherein, the target cold storage capacity The calculation strategy is: ； in, is the photovoltaic power generation power at time t; is the base load power at time t; It is the latent heat absorbed or released during the phase change of building materials; S55: Evaluate the mismatch degree from three different dimensions: photovoltaic output, air conditioning load, and thermal capacity, and obtain photovoltaic evaluation items, air conditioning load evaluation items, and thermal capacity matching items. Then, simultaneously evaluate the photovoltaic-air conditioning load matching degree from these three dimensions. .

8. A large model-based dynamic energy management method for a central air-conditioning system according to claim 7, characterized in that: The S5 also includes: S56: Extracting PV-air conditioning load matching , to switch the central air-conditioning working mode, including: Preset matching threshold , and conduct synergy level analysis based on the matching degree. When the PV-air conditioning load matching degree is Greater than or equal to When , the collaborative control execution mode is triggered; When the PV-air conditioning load matching degree Less than When , the coordinated control execution unit is not triggered and the backup power grid power supply strategy is started; The collaborative control execution mode includes: increasing the operating power of the air conditioning system during peak photovoltaic output periods, storing excess cooling energy in the building structure, and releasing the stored cooling energy during peak electricity price periods.

9. A large-scale model-based dynamic energy management system for a central air-conditioning system, used to implement a large-scale model-based dynamic energy management method for a central air-conditioning system as claimed in any one of claims 1 to 8, characterized in that: The system comprises: Load forecasting module, thermal inertia quantification module, phase variable quantification module, dynamic energy storage control module and air conditioning operation optimization module; The load forecasting module is used to integrate real-time indoor and outdoor environmental data and central air-conditioning equipment status data, and output short-term load demand forecast curves through a time series deep prediction model; The thermal inertia quantification module is used to collect temperature field data of building materials, calculate the equivalent heat capacity and equivalent thermal resistance of building materials through physical information neural network inversion, and quantify the thermal inertia parameters of building materials in real time; The phase change quantification module is used to predict the temperature field change trend of the building structure in the next three hours, and at the same time, based on the predicted temperature field change trend, evaluate the phase change delay time coefficient of the building materials; The dynamic energy storage control module is used to obtain the photovoltaic power generation power prediction curve of the power grid system and the power grid time-sharing electricity price signal to determine the optimal charging and discharging time window for building materials; The air conditioning operation optimization module adopts a multi-objective decision-making algorithm to generate the optimal operation strategy of the central air conditioning system.

Images