An Optimized Scheduling Method for Building Energy Systems Based on Efficient Utilization of Active and Passive Energy Storage

Abstract

This invention provides an optimized scheduling method for building energy systems based on efficient utilization of both active and passive energy storage, belonging to the field of energy management and intelligent control technology. By integrating multiple flexible resources, it achieves efficient energy utilization, cost optimization, and guaranteed thermal comfort. A four-terminal prediction model is established, combining weather forecasts and operational data to accurately predict energy supply and demand for the next 24 hours. A two-stage optimized scheduling approach is adopted: the first stage optimizes the heating water temperature of the heat pump, and the second stage refines the scheduling of energy storage batteries and other equipment. This method achieves multi-objective collaborative optimization, fully taps the potential of flexible resources, reduces computational complexity, facilitates practical application, effectively improves energy utilization efficiency, reduces operating costs, ensures thermal comfort, and promotes the intelligent and efficient development of building energy management systems, demonstrating significant economic and social benefits.

Description

Technical Field

[0001] This invention relates to the field of energy management and intelligent control technology, and in particular to an optimized scheduling method for building energy systems based on the efficient utilization of active and passive energy storage. Background Technology

[0002] With the rapid development of renewable energy technologies, energy devices such as photovoltaic power generation, air source heat pumps, energy storage batteries, and smart home appliances are increasingly being used in buildings. However, existing building energy systems still face many problems in energy management, particularly the mismatch between photovoltaic power generation and the system's electricity supply and demand. Specifically, these problems manifest as follows:

[0003] (1) Lack of coordinated optimization scheduling of multiple flexible resources: Existing systems usually manage flexible resource devices such as photovoltaic power generation, air source heat pumps, energy storage batteries, and reverse charging piles independently, lacking a unified scheduling optimization strategy, making it difficult to achieve efficient energy utilization and cost minimization.

[0004] (2) Insufficient coordination between heat pump system scheduling and grid and photovoltaic power generation: Traditional heat pump system operation strategies mostly consider unit efficiency and user thermal comfort experience, but fail to fully consider the volatility of photovoltaic power generation and the impact of time-of-use electricity pricing, leading to energy waste or increased operating costs. For example, during peak photovoltaic power generation periods, the heat pump may not fully utilize photovoltaic power generation, while during peak electricity consumption periods, the heat pump may increase the burden on the grid. Therefore, in the process of multi-objective optimization, there may be a contradiction between unit efficiency, cost-saving goals and user experience.

[0005] (3) Single energy storage battery scheduling strategy: Existing energy storage systems and electric vehicles mostly adopt fixed and user-defined charging and discharging modes, failing to dynamically optimize scheduling based on photovoltaic power generation, time-of-use electricity price and user demand, resulting in low utilization of energy storage batteries, and may even shorten their lifespan due to overcharging and discharging.

[0006] (4) Lack of intelligent scheduling of transferable load appliances: The scheduling of transferable load appliances such as washing machines and water heaters is usually based on fixed time or manual control by users. It fails to combine photovoltaic power generation, time-of-use pricing and user living habits for intelligent optimization, which reduces energy utilization efficiency.

[0007] (5) Inefficient scheduling algorithms: Existing optimization scheduling methods typically employ multi-parameter centralized solution algorithms, simultaneously optimizing various flexible resources. In practical engineering applications, this method suffers from high computational complexity and low solution efficiency, placing high demands on computing power and making it difficult to meet the real-time and practical requirements of building energy management systems.

[0008] Therefore, there is an urgent need for a building energy management system that can coordinate and optimize the use of multiple flexible resources such as active and passive energy storage, take into account both thermal comfort and energy-saving goals, and achieve high efficiency, so as to improve energy utilization efficiency, reduce operating costs and ensure thermal comfort. Summary of the Invention

[0009] The purpose of this invention is to provide an optimized scheduling method for building energy systems based on the efficient utilization of active and passive energy storage. This method optimizes the scheduling of building energy systems integrating photovoltaic power generation, air-source heat pumps, energy storage batteries, reverse-charging trolley piles, adjustable lighting fixtures, and smart home appliances (such as washing machines and water heaters) to achieve efficient energy utilization, cost optimization, and user thermal comfort. By using weather forecasts and building energy system operation data, a four-terminal prediction model is established, encompassing the building's dynamic thermal environment, photovoltaic power generation, appliance load, and heat pump energy consumption. This model accurately predicts both supply and demand for the energy system over the next 24 hours. Combined with a two-stage optimized scheduling method, the first stage focuses on optimizing the heating water temperature of the air-source heat pump, integrating photovoltaic power generation and time-of-use pricing to ensure thermal comfort. The second stage, based on the remaining photovoltaic power generation capacity, performs refined scheduling of energy storage batteries, reverse-charging piles, adjustable lighting fixtures, and transferable load appliances. This achieves efficient utilization of various flexible resources within the energy system, including building thermal inertia, active energy storage, and adjustable home appliances, ultimately maximizing energy efficiency, ensuring thermal comfort, and optimizing economic performance.

[0010] To achieve the above objectives, this invention provides an optimized scheduling method for building energy systems based on efficient utilization of active and passive energy storage, comprising the following steps:

[0011] Step S1: Collect building thermal information, heating system operation data, photovoltaic power generation system operation data, home appliance operation data, electric vehicle and energy storage system information;

[0012] Step S2: Establish four-terminal prediction models, including: building heat transfer dynamic prediction model, heat pump energy consumption prediction model, photovoltaic power generation prediction model, and household appliance load prediction model;

[0013] Step S3: Obtain the weather forecast data for the next 24 hours and the time-of-use electricity price;

[0014] Step S4: Input the data obtained in step S3 into the model in step S2 to obtain the predicted values ​​of indoor temperature, heat load, heat pump energy consumption, photovoltaic power generation and household appliance load for the next 24 hours.

[0015] Step S5: Perform two-stage optimized scheduling.

[0016] Preferably, in step S1, the building thermal information includes the building envelope dimensions, materials, and heat transfer coefficient; the heating system operation data includes the heating system water temperature, outdoor air temperature, solar radiation intensity, indoor air temperature, and energy consumption; the photovoltaic power generation system operation data includes outdoor air temperature, solar radiation intensity, outdoor air humidity, cloud cover, and photovoltaic power generation power; the home appliance operation data includes the type of home appliance, the range of home appliance usage time, adjustable power, and total electrical load of home appliances; and the electric vehicle and energy storage system information includes battery capacity, depth of charge and discharge, range of charge and discharge power, charge and discharge efficiency, and initial state of charge.

[0017] Preferably, in step S2, a neural network is used to establish a heat pump energy consumption prediction model, with the input parameters being outdoor temperature, supply water temperature and return water temperature, and the output parameter being unit energy consumption.

[0018] A photovoltaic power generation prediction model is established using a neural network. The input parameters are solar radiation intensity, outdoor temperature, humidity, and cloud cover, and the output parameter is photovoltaic power generation.

[0019] A neural network was used to establish a home appliance load prediction model, with outdoor temperature and humidity and historical electrical load as input parameters.

[0020] Preferably, in step S2, a dynamic prediction model for building heat transfer is constructed based on the principle of thermal resistance and thermal capacity analogy, as follows:

[0021]

[0022] Among them, C ai Indicates the heat capacity of indoor air, T ai T represents the predicted indoor air temperature. w_in R represents the inner surface temperature of a non-transparent enclosure structure. w_in T represents the thermal resistance between the inner surface of a non-transparent building envelope and the indoor air. r_in R represents the temperature of the inner surface of a non-transparent roof envelope. r_in T represents the thermal resistance of a non-transparent building envelope roof to the indoor air. im R represents the temperature of the internal heat storage element. im T represents the thermal resistance between the internal heat storage element and the indoor air. f R represents the radiant floor temperature. f Q represents the thermal resistance between the radiant floor and the indoor air. p Q represents the heat dissipation of personnel / home appliances. win Indicates heat transfer in a transparent enclosure structure, C w and C r T represents the heat capacity of the non-transparent building envelope walls and roof, respectively. w_out and T r_out T represents the outer surface temperature of the non-transparent building envelope walls and roof, respectively.w_in and T r_in R represents the inner surface temperature of the non-transparent enclosure walls and roof, respectively. w and R r R represents the thermal resistance of the non-transparent building envelope walls and roof, respectively. w_out and R r_out R represents the thermal resistance of the non-transparent building envelope walls and roof exterior surfaces to the outside air, respectively. w_in and R r_in These represent the thermal resistance of the non-transparent building envelope walls and the inner surface of the roof to the indoor air, respectively; T ao Q represents the outdoor air temperature. solar_w and Q solar_r C represents the solar radiation heat gain of the non-transparent building envelope walls and roof surface, respectively. im Q represents the heat capacity of the internal heat storage body. solar_im This indicates the heat gained through solar radiation that is exchanged between the window and the internal heat storage unit.

[0023] Preferably, in step S5, the first stage aims to minimize system operating costs and indoor temperature deviation. A biomimetic algorithm is used to dynamically optimize the air-source heat pump water supply temperature, combined with building thermal inertia to achieve passive energy storage scheduling. The objective function for the first stage of optimization scheduling is:

[0024] J(T ws,set,0 ,T ws,set,2 ,...,T ws,set,23 )=Min(a×C HVAC +ΔT ai );

[0025]

[0026] Among them, J(T) ws,set,0 ,T ws,set,2 ,...,T ws,set,23 T represents the objective function value for water temperature regulation. ws,set This represents the optimal water temperature setpoint for the air source heat pump, and 'a' represents the cost weighting factor, used to adjust the importance of heating system operating costs in the objective function. and These represent the minimum and maximum allowable values ​​for the water temperature setting, ΔT and ΔT, respectively. ai C is the root mean square deviation of room temperature, used to characterize indoor thermal comfort. HVAC The formula for calculating the operating cost of the heating system for the next day is as follows:

[0027] P HVAC (t)=P ASHP (t)+P PUMP (t);

[0028]

[0029] Among them, P HVAC (t) represents the system power consumption, which is determined by the heat pump power consumption P. ASHP (t) and water pump power consumption P PUMP (t) constitutes, P HVAC_grid (t) represents the electricity consumption of the heat pump system from the power grid, P PV (t) represents the photovoltaic power generation at time t, y TOU (t) represents the time-of-use electricity price at time t;

[0030] ΔT ai The root mean square deviation of room temperature is calculated based on the hourly indoor temperature T for the next day. ai (t) and the hourly indoor set temperature T for the next day ai,set (t) is calculated to yield:

[0031]

[0032] Preferably, in step S5, in the second stage, based on the unused photovoltaic power generation from the first stage, a mixed-integer linear programming algorithm is used to optimize the charging and discharging power of energy storage batteries, the charging and discharging power of electric vehicles, and the operating power of adjustable home appliances. The objective function for the second stage of optimization scheduling is:

[0033] J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j1,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 ) = Min(C sys );

[0034] Among them, J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 P represents the objective function for the two-stage building energy system scheduling. c P dc P represents the optimal battery charging and discharging power. ev_c P ev_dcP represents the optimal charging and discharging power of an electric vehicle. ad,z P represents the power consumption of the z-th adjustable home appliance. sh,j This represents the power consumption of the household appliance with the j-th type of transferred load;

[0035] C sys The formula for calculating the building's energy system operating cost for the next day is as follows:

[0036]

[0037] Among them, P grid p(t) represents the amount of electricity that needs to be purchased from the grid after optimized scheduling, and p(t) represents the time-of-use electricity price at time t. The solution process establishes a power balance model as follows:

[0038] P grid (t)+P dc (t)+P ev_dc (t)≥P HVAC (t);

[0039] P grid (t)+P dc (t)+P ev_dc (t)+P pv_unused (t)≥P app (t)+P app_fle (t)+P c (t)+P ev_c (t);

[0040] Among them, P pv_unused (t) represents the photovoltaic power generation that was not absorbed in the first stage, P c (t), P ev_c (t) and P dc (t), P ev_dc (t) represents the charging and discharging power of the battery and the electric vehicle, P app (t) and P app_fle (t) represents the basic load and flexible load of the remaining household appliances, P app (t) is derived from the prediction model.

[0041] Preferred, P app_fle The solution of (t) is mainly divided into two types: flexibility of power-adjustable equipment and flexibility of load-transferable equipment.

[0042] P app_fle (t)=P app_fle,ad (t)+P app_fle,sh (t);

[0043]

[0044] twork,s ≥t window,s t work,e ≤t window,e l work ≤l window ;

[0045]

[0046] Where, N z (t) represents the number of adjustable-power home appliances of the z-th type at time t, M j (t) represents the number of the j-th type of load-transferable home appliance that can be adjusted at time t, P ad,z (t) represents the power that the z-th type of power-adjustable home appliance can reduce at time t, P sh,j (t) represents the operating power of the j-th load-transferable household appliance, U j (t) represents the flexible state of the j-th load-transferable appliance at time t, n represents the types of power-adjustable appliances that can be adjusted at time t, m represents the types of load-transferable appliances that can be adjusted at time t, and P app_fle,ad (t) represents the flexible load that the adjustable home appliance can provide at time t, P app_fle,sh (t) represents the flexible load that the portable appliance can provide at time t, l work,shift t represents the transferable time period within the work window. work,s Indicates the start time of equipment operation, t window,s Indicates the start time of the time window, t work,e t represents the end time of the equipment's operation. window,e Indicates the end time of the time window, l work Indicates working hours, l window,s Indicates the duration of the time window;

[0047] The formula for calculating flexibility is as follows:

[0048] Q app_fle (t)=Q app_fle,ad (t)+Q app_fle,sh (t);

[0049]

[0050] Where, n z This represents the heat dissipation coefficient of the z-th type of adjustable power home appliance. C represents the cooling load coefficient for the heat dissipation of the z-th type of adjustable power household appliance. sh,jLet n represent the operating power of the j-th load-transferable household appliance, n1 represent the ballast power consumption coefficient of the fluorescent lamp, n2 represent the lampshade heat insulation coefficient of the fluorescent lamp, n3 represent the motor utilization coefficient, n4 represent the motor load coefficient, n5 represent the heat dissipation coefficient of the electric heating equipment or electronic equipment, η represent the motor efficiency, and Q represent the load coefficient of the electric motor. app_fle (t) represents the total flexible load of the remaining household appliances due to heat dissipation at time t, Q. app_fle,ad (t) represents the total flexible load that a power-adjustable household appliance can provide due to heat dissipation at time t, Q. app_fle,sh (t) represents the total flexible load that a power-transferable household appliance can provide at time t due to heat dissipation, and COP represents the performance coefficient of the household appliance.

[0051] Preferably, the battery's charging and discharging power P c (t) and P dc(t) Subject to the following battery scheduling process constraints:

[0052]

[0053] Among them, SOC (t) The SOC(t0) represents the current energy state of the battery, and the SOC(t0) represents the initial energy state of the battery at time t0. η c and η dc E represents the battery charging and discharging efficiency, Δt represents the time interval, and E Max This refers to the battery's rated capacity.

[0054] Batteries are subject to maximum and minimum capacity constraints during scheduling, specifically:

[0055] SOC min ≤SOC(t)≤SOC max ;

[0056] Among them, SOC max and SOC min Indicates the battery's maximum and minimum capacity;

[0057] During the scheduling process, the battery is subject to the maximum and minimum allowable charging and discharging power constraints, specifically:

[0058]

[0059] in, and This indicates the battery's maximum charge and discharge power;

[0060] Since batteries cannot be charged and discharged simultaneously, binary variables are introduced to enforce mutual exclusion constraints on charge and discharge states:

[0061] x c +xdc ≤1;

[0062] Where, x c and x dc Indicates the battery's charge / discharge state as 0 and 1;

[0063] Charge-discharge cycle constraints:

[0064]

[0065] Where T represents the time range covered by the optimized scheduling.

[0066] Therefore, this invention employs the aforementioned method for optimizing and scheduling building energy systems based on efficient utilization of active and passive energy storage. By integrating various flexible resources such as photovoltaic power generation, air source heat pumps, energy storage batteries, reverse trolley charging stations, adjustable lighting fixtures, and smart home appliances (e.g., washing machines, water heaters), it achieves multi-objective optimization of efficient energy utilization, cost optimization, and user thermal comfort assurance. The beneficial technical effects are as follows:

[0067] This invention establishes a four-terminal prediction model, combining weather forecasts and building energy system operation data, to achieve accurate predictions of both supply and demand in the energy system for the next 24 hours. This prediction model fully considers building thermal inertia, photovoltaic power generation volatility, and changes in heat pump energy consumption, significantly improving prediction accuracy and enhancing the robustness of scheduling strategies.

[0068] This invention employs a two-stage optimization scheduling method. The first stage optimizes the heating water temperature of the air-source heat pump, combined with photovoltaic power generation and time-of-use pricing, to ensure indoor thermal comfort. The second stage, based on the remaining photovoltaic power generation, performs refined scheduling of energy storage batteries, reverse charging piles, adjustable lighting fixtures, and transferable load appliances. This method achieves synergistic optimization of multiple objectives, including economy, comfort, and energy efficiency, solving the energy waste and cost increases caused by traditional single scheduling strategies, and fully tapping the potential of various flexible resources such as building thermal inertia, active energy storage, and adjustable appliances.

[0069] This invention employs a two-stage optimization scheduling method, decomposing the complex problem into two stages: heat pump water temperature scheduling and flexible resource scheduling, respectively solving for the system control parameters under multi-objective optimization. This method effectively reduces computational complexity, decreases dependence on computing power, and facilitates rapid deployment and application in practical engineering. Attached Figure Description

[0070] Figure 1 A flowchart of an optimized scheduling method for building energy systems based on efficient utilization of active and passive energy storage;

[0071] Figure 2 Diagram of a dynamic prediction model for building heat transfer;

[0072] Figure 3 This relates the working time of home appliances to the time window. Detailed Implementation

[0073] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0074] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0075] Example 1

[0076] This invention provides an optimized scheduling method for building energy systems based on efficient utilization of active and passive energy storage. Based on indoor temperature and heat load prediction, air source heat pump energy consumption prediction, photovoltaic power generation prediction, and prediction of other household appliance electricity loads, it further dynamically plans the control parameters of the air source heat pump heating water temperature setpoint, battery and electric vehicle charging and discharging behavior, and other flexible resources. While ensuring indoor thermal comfort, it responds to 24-hour time-of-use electricity price changes and photovoltaic power generation, alleviating the energy supply-demand mismatch problem and reducing system operating costs. Specifically, it includes the following steps:

[0077] Step S1: Collect data, including:

[0078] (1) Building thermal information: building envelope dimensions, materials and heat transfer coefficient;

[0079] (2) Heating system operation data: heating system water temperature, outdoor air temperature, solar radiation intensity, indoor air temperature and energy consumption;

[0080] (3) Photovoltaic power generation system operation data: outdoor air temperature, solar radiation intensity, outdoor air humidity, cloud cover and photovoltaic power generation power;

[0081] (4) Home appliance operation data: types of home appliances, range of home appliance usage time, adjustable power and total electrical load of home appliances;

[0082] (5) Electric vehicle and energy storage system information: battery capacity, depth of charge and discharge, charge and discharge power range, charge and discharge efficiency and initial state of charge (SOC).

[0083] Step S2: Based on building thermal information and heating system operation data, predict the changes in indoor temperature and heating load over the next 24 hours under different outdoor environments and different water supply temperatures to ensure that thermal comfort is not affected during scheduling. This invention is based on the principle of thermal resistance and thermal capacity analogy, and considers the thermal inertia brought about by outdoor air temperature, solar radiation intensity, water supply temperature, building envelope, internal heat storage body, radiant coils and radiant floor, etc., to establish a building dynamic thermal environment prediction model, which aims to predict dynamic indoor temperature changes as shown in formulas (1)-(6).

[0084]

[0085] Among them, C ai Indicates the heat capacity of indoor air, T ai T represents the predicted indoor air temperature. w_in R represents the inner surface temperature of a non-transparent enclosure structure. w_in T represents the thermal resistance between the inner surface of a non-transparent building envelope and the indoor air. r_in R represents the temperature of the inner surface of a non-transparent roof envelope. r_in T represents the thermal resistance of a non-transparent building envelope roof to the indoor air. im R represents the temperature of the internal heat storage element. im T represents the thermal resistance between the internal heat storage element and the indoor air. f R represents the radiant floor temperature. f Q represents the thermal resistance between the radiant floor and the indoor air. p Q represents the heat dissipation of personnel / home appliances. win This indicates heat transfer through a transparent enclosure structure.

[0086] Among them, for the heat transfer of the room through the non-transparent enclosure structure in the first term on the right side of formula (1), the heat transfer of the wall is further calculated as shown in formula (2) and formula (3), and the heat transfer of the roof is further calculated as shown in formula (4) and formula (5).

[0087]

[0088] Among them, C w and C r T represents the heat capacity of the non-transparent building envelope walls and roof, respectively. w_out and T r_out T represents the outer surface temperature of the non-transparent building envelope walls and roof, respectively. w_in and T r_in R represents the inner surface temperature of the non-transparent enclosure walls and roof, respectively. w and R r R represents the thermal resistance of the non-transparent building envelope walls and roof, respectively. w_out and R r_outR represents the thermal resistance of the non-transparent building envelope walls and roof exterior surfaces to the outside air, respectively. w_in and R r_in T represents the thermal resistance of the inner surfaces of the non-transparent building envelope walls and roof to the indoor air, respectively. ao Q represents the outdoor air temperature. solar_w and Q solar_r These represent the heat gain from solar radiation on the non-transparent building envelope walls and roof surface, respectively.

[0089] The heat storage body inside the building, composed of interior walls, furniture, etc., continuously increases its temperature through convection or solar radiation through windows. When the surface temperature of the heat storage body is higher than the indoor air temperature, the accumulated heat is gradually released into the indoor air through convection, ensuring indoor comfort.

[0090] The calculation method for the heat storage and release characteristics of the heat storage body inside the building in formula (1) is shown in formula (6).

[0091]

[0092] Among them, C im Q represents the heat capacity of the internal heat storage body. solar_im This indicates the heat gained through solar radiation that is exchanged between the window and the internal heat storage unit.

[0093] The heat gain from solar radiation can be calculated using formula (7).

[0094] Q solar =v×I(7);

[0095] Among them, Q solar The value represents the heat of solar radiation, v represents the solar radiation absorption coefficient, which is affected by factors such as absorptivity, orientation, or angle, and I represents the predicted value of solar radiation intensity, obtained from a meteorological website.

[0096] For heat transfer Q of transparent enclosure structures win The calculation method is shown in formula (8).

[0097]

[0098] Among them, R win The thermal resistance is for the transparent enclosure structure.

[0099] The heat pump water system, heating terminal and indoor heat exchange are calculated as shown in formula (9).

[0100]

[0101] Among them, C f T represents the floor heat capacity. ws This indicates the water supply temperature of the heat pump.

[0102] Since some parameters in the model are affected by changes in the outdoor environment, such as increased wind speed leading to a decrease in surface thermal resistance, in order to obtain accurate values ​​for each heat capacity, thermal resistance, and solar radiation coefficient, a rolling optimization method using biomimetic algorithms or least squares can be used. The genetic algorithm input parameter set consists of the model prediction and actual measurement values ​​of the indoor air temperature of the heating system for the most recent 5 consecutive days with a time interval of 1 hour. The goal is to minimize the deviation between the model-predicted indoor temperature and the actual indoor temperature. The rolling optimization identifies the values ​​of each parameter at a frequency of 1 time per day. The specific objective function F(.) is shown in formula (10).

[0103]

[0104] Where t1 and t2 are the start and end times of the identification and optimization process, respectively, with an interval of 120 hours. ai,test (t) represents the indoor air temperature monitoring value.

[0105] After establishing and optimizing the building heat transfer dynamic prediction model, the outdoor air temperature, solar radiation, and air source heat pump water supply temperature for the next 24 hours can be input into the model to predict the indoor air temperature for the next 24 hours. The model structure is as follows: Figure 2 As shown.

[0106] The system heating load depends primarily on the supply water temperature and the radiant floor temperature. Since the radiant floor has a large heat capacity, its temperature fluctuations are relatively stable. Therefore, this invention uses the radiant floor temperature obtained from the previous iteration and the current supply water temperature to calculate the system heating load, as shown in formula (11):

[0107]

[0108] Among them, Q ws R represents the system heating load. pf This indicates the thermal resistance between the radiant floor terminal and the indoor air.

[0109] Accurately predicting the energy consumption of air source heat pumps, photovoltaic power generation, and the load of other household appliances in the next 24 hours under different outdoor environments and heating water temperatures is the basic data support for optimizing scheduling. Using machine learning algorithms, photovoltaic power generation prediction models, household appliance load prediction models, and heat pump energy consumption prediction models are established to complete the prediction of both supply and demand.

[0110] The actual heating performance of air source heat pumps is affected by the supply water temperature and outdoor operating conditions. This invention is based on actual operating data of air source heat pumps, using outdoor temperature (T) as an example. ao ), water supply temperature (T) ws ) and return water temperature (T wr) as input parameters, with unit energy consumption (P) ASHP Using ) as the output parameter, a heat pump energy consumption prediction model is established using neural network algorithms such as CNN or ANN.

[0111] Distributed photovoltaic power generation is mainly affected by outdoor environmental conditions. This invention selects solar radiation intensity (I) and outdoor temperature (T) that can be obtained from weather forecasts. ao Outdoor humidity (RH) and cloud cover (Cc) are used as input parameters, and photovoltaic power generation (P) is used as the input parameter. pv Using ) as the output parameter, a photovoltaic power generation prediction model is established using neural network algorithms such as CNN or LSTM.

[0112] The energy consumption of other household appliances (such as lighting, televisions, and refrigerators) within the building is mainly influenced by user behavior patterns, environmental parameters, and user characteristics. To construct an accurate energy consumption prediction model, this method first uses time characteristics (different seasons, weekdays, and holidays) to cluster historical data. Secondly, it selects the daily electricity load (P... t-24 ), outdoor temperature (T) ao Outdoor humidity (RH) is used as the input parameter, and the total load of household appliances (P) is used as the input parameter. app Using ) as the output parameter, a home appliance load prediction model is established using neural network algorithms such as CNN or LSTM.

[0113] Step S3: Collect future weather data and local power grid electricity prices. Specific data include outdoor air temperature, solar radiation intensity, heating water temperature, indoor temperature, and time-of-use electricity prices.

[0114] Step S4: Collect future meteorological data (including solar radiation intensity (I) and outdoor temperature (T)) from step S3. ao The outdoor humidity (RH) and cloud cover (Cc) and time-of-use electricity price information are respectively input into the building dynamic thermal environment prediction model, photovoltaic power generation prediction model, household appliance load prediction model and heat pump energy consumption prediction model established in step S2, to obtain the indoor temperature change, indoor heat load, photovoltaic power generation power and air source heat pump heating system energy consumption under different operating conditions and different heating water temperature settings.

[0115] Step S5: With the goal of minimizing the deviation between the system operating cost and the set indoor temperature over the next 24 hours, dynamically plan the air source heat pump heating water temperature using particle swarm optimization or other biomimetic algorithms, and combine this with the MILP algorithm to plan the charging and discharging behavior of the battery and electric vehicle, as well as the electricity consumption behavior of household appliances. The optimization scheduling is divided into two stages:

[0116] (1) The first stage (constructing the target optimization function of water temperature) is to schedule the air source heat pump in combination with the building's thermal inertia passive energy storage. The main adjustment parameter is the air source heat pump water supply temperature. The goal of the heating water temperature optimization scheduling is to maximize the response to electricity demand, absorb photovoltaic power generation, and reduce operating costs under the premise of ensuring indoor thermal comfort. It is described by formulas (12) to (13).

[0117] J(T ws,set,0 ,T ws,set,2 ,...,T ws,set,23 )=Min(a×C HVAC +ΔT ai (12);

[0118]

[0119] Among them, J(T) ws,set,0 ,T ws,set,2 ,...,T ws,set,23 T represents the objective function value for water temperature regulation. ws,set This represents the optimal water temperature setpoint for the air source heat pump, and 'a' represents the cost weighting factor, used to adjust the importance of heating system operating costs in the objective function. and These represent the minimum and maximum allowable values ​​for the water temperature setting, ΔT and ΔT, respectively. ai is the root mean square deviation of room temperature, used to characterize indoor thermal comfort.

[0120] C HVAC This represents the operating cost of the heating system for the next day, calculated according to formulas (14)-(16). Wherein, P... HVAC (t) represents the system power consumption, which is determined by the heat pump power consumption P. ASHP (t) and water pump power consumption P PUMP (t) constitutes, P HVAC_grid (t) represents the grid power consumption of the heat pump system. When the system power consumption is lower than the photovoltaic power generation, grid power is not required; conversely, grid power is needed to supplement the insufficient photovoltaic power consumption of the system. C HVAC It is calculated based on the power grid's electricity consumption and electricity costs.

[0121] P HVAC (t)=P ASHP (t)+P PUMP (t)(14);

[0122]

[0123] Among them, P PV (t) represents the photovoltaic power generation at time t, y TOU (t) represents the time-of-use electricity price at time t.

[0124] ΔT ai The root mean square deviation of room temperature is used to characterize indoor thermal comfort, based on the hourly indoor temperature T for the next day. ai (t) and the hourly indoor set temperature T for the next day ai,set (t) is calculated as shown in formula (17).

[0125]

[0126] (2) The second stage (energy system optimization objective function) is to schedule the battery, electric vehicle and other household appliance loads. The main adjustment parameters are the charging and discharging power of the battery and electric vehicle and the power consumption of the other household appliances. Based on the optimization results of the first stage, the photovoltaic power generation is further absorbed to reduce the operating cost of the entire building system. It is described by formula (18).

[0127] J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j1,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 ) = Min(C sys (18);

[0128] Among them, J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 P represents the objective function for the two-stage building energy system scheduling. c P dc P represents the optimal battery charging and discharging power. ev_c P ev_dc P represents the optimal charging and discharging power of an electric vehicle. ad,z P represents the power consumption of the z-th adjustable home appliance. sh,j This represents the power consumption of the household appliance with the j-th type of transferred load.

[0129] C sys The formula for calculating the building's energy system operating cost for the next day is as follows:

[0130]

[0131] Among them, P grid p(t) represents the amount of electricity that needs to be purchased from the grid after optimized scheduling, and p(t) represents the time-of-use electricity price at time t. The solution process requires establishing a power balance model, as follows:

[0132] P grid (t)+P dc (t)+P ev_dc (t)≥P HVAC (t)(20);

[0133] P grid (t)+P dc (t)+P ev_dc (t)+P pv_unused (t)≥P app (t)+P app_fle (t)+P c (t)+P ev_c (t) (21);

[0134] Among them, P pv_unused (t) represents the photovoltaic power generation that was not absorbed in the first stage, P c (t), P ev_c (t) and P dc (t), P ev_dc (t) represents the charging and discharging power of the battery and the electric vehicle, P app (t) and P app_fle (t) represents the basic load and flexible load of the remaining household appliances, P app (t) is derived from the prediction model.

[0135] P app_fle The solution to (t) mainly falls into two categories: flexibility for power-adjustable devices and flexibility for load-transferable devices. Power-adjustable devices provide flexibility by adjusting their power output, such as adjusting the power of lighting equipment. Load-transferable devices can complete tasks at any time within a time window without affecting their functionality, thus providing power flexibility; examples include washing machines, dishwashers, and coffee machines in buildings. A time window refers to the acceptable operating period for an appliance, while the operating duration refers to the actual time required for the device to operate. Generally, both the time window and the required operating duration for each appliance are constant values, and the duration of the time window is l. window,s Not less than working hours l work The start time t of the equipment operation work,s No earlier than the start time t of the time window window,s The end time t of the equipment operation work,e No later than the end time t of the time window window,eA diagram illustrating the relationship between working hours and time windows is shown below. Figure 3 As shown.

[0136] Based on the correspondence between time windows and working hours, the flexibility quantification model of home appliances is defined by formula (21), where home appliance flexibility includes the flexibility of power-adjustable equipment and the flexibility of load-transferable equipment, and the calculation formulas are shown in formulas (22) and (23), respectively. The flexibility of power-adjustable equipment mainly comes from various lighting equipment in buildings. During the demand response period, the power of such equipment can be reduced by a certain proportion, thereby reducing the total electricity load of the building. The flexibility of load-transferable equipment mainly comes from home appliances whose working hours can be changed. By changing the working time of the equipment within the time window, the purpose of load transfer can be achieved, which will not have a significant impact on the building and users, and will not require additional investment and operating costs.

[0137] P app_fle (t)=P app_fle,ad (t)+P app_fle,sh (t)(22);

[0138]

[0139] t work,s ≥t window,s t work,e ≤t window,e l work ≤l window (25);

[0140]

[0141] Where, N z (t) represents the number of adjustable-power home appliances of the z-th type at time t, M j (t) represents the number of the j-th type of load-transferable home appliance that can be adjusted at time t, P ad,z (t) represents the power that the z-th type of power-adjustable home appliance can reduce at time t, P sh,j (t) represents the operating power of the j-th load-transferable household appliance, U j (t) represents the flexible state of the j-th load-transferable appliance at time t, n represents the types of power-adjustable appliances that can be adjusted at time t, m represents the types of load-transferable appliances that can be adjusted at time t, and P app_fle,ad (t) represents the flexible load that the adjustable home appliance can provide at time t, P app_fle,sh (t) represents the flexible load that the portable appliance can provide at time t, l work,shift This indicates the time period that can be moved within the work window.

[0142] For power-adjustable and load-transferable appliances, when their participation in demand response brings direct power flexibility, changes in appliance power and load transfer will cause changes in the appliance's heat dissipation, indirectly causing fluctuations in heating and cooling loads. Ultimately, these fluctuations will alter the energy consumption of the air conditioning system, further leading to changes in the electrical load. Therefore, appliance heat dissipation also possesses a degree of power flexibility. The flexibility resulting from changes in appliance heat dissipation can be divided into heat dissipation flexibility for power-adjustable appliances and heat dissipation flexibility for load-transferable appliances. The formula for calculating this flexibility is as follows:

[0143] Q app_fle (t)=Q app_fle,ad (t)+Q app_fle,sh (t)(27);

[0144]

[0145] Where, n z This represents the heat dissipation coefficient of the z-th type of adjustable power home appliance. C represents the cooling load coefficient for the heat dissipation of the z-th type of adjustable power household appliance. sh,j Let n represent the operating power of the j-th type of load-transferable household appliance, n1 represent the power consumption coefficient of the fluorescent lamp ballast (generally taken as 1.2), n2 represent the heat insulation coefficient of the fluorescent lamp shade (0.5-0.6 for perforated shades, 0.6-0.8 for non-perforated shades), n3 represent the motor utilization coefficient (ratio of maximum actual power to installed power of the motor, generally taken as 0.7-0.9), n4 represent the motor load coefficient (ratio of average actual power to maximum actual power of the motor, generally taken as 0.15-0.5), n5 represent the heat dissipation coefficient of electric heating equipment or electronic equipment (1 for electric fans, computers, etc., 0.5-0.9 for other equipment), η represent the motor efficiency (1 for electric fans, computers, etc., 0.5-0.9 for other equipment), and Q represent the motor efficiency. app_fle (t) represents the total flexible load of the remaining household appliances due to heat dissipation at time t, Q. app_fle,ad (t) represents the total flexible load that a power-adjustable household appliance can provide due to heat dissipation at time t, Q. app_fle,sh (t) represents the total flexible load that a power-transferable household appliance can provide at time t due to heat dissipation, and COP (Coefficient of Performance) represents the performance coefficient of the household appliance.

[0146] Secondly, the battery's charging and discharging power P c (t) and P dc(t) Subject to the following battery scheduling process constraints:

[0147]

[0148] Among them, SOC (t)The SOC(t0) represents the current energy state of the battery, and the SOC(t0) represents the initial energy state of the battery at time t0. η c and η dc This represents the battery's charging and discharging efficiency, typically a value between 0 and 1. It indicates the battery's energy conversion efficiency during the charging process, with Δt representing the time interval. Max This refers to the battery's rated capacity.

[0149] Batteries are subject to maximum and minimum capacity constraints during scheduling, specifically:

[0150] SOC min ≤SOC(t)≤SOC max (32);

[0151] Among them, SOC max and SOC min This indicates the battery's maximum and minimum capacity.

[0152] During the scheduling process, the battery is subject to the maximum and minimum allowable charging and discharging power constraints, specifically:

[0153]

[0154] in, and This indicates the battery's maximum charge and discharge power.

[0155] Since batteries cannot be charged and discharged simultaneously, binary variables are introduced to enforce mutual exclusion constraints on charge and discharge states:

[0156] x c +x dc ≤1(35);

[0157] Where, x c and x dc The numbers 0 and 1 represent the battery's charge / discharge state, ensuring that the battery can only be in a charging or discharging state at any given time.

[0158] To ensure battery health, charge / discharge cycle constraints are set:

[0159]

[0160] Where T represents the time range covered by the optimized scheduling, which is 24 hours in this embodiment.

[0161] The constraints and scheduling model for electric vehicles are the same as those for batteries, but they need to be charged and discharged within a fixed time window.

[0162] Taking a residential building in a certain city as an example, and in conjunction with the accompanying drawings, the specific implementation method of the present invention will be further described in detail.

[0163] The residential building area is 198m² 2 The exterior walls are made of polyurethane insulated panels with double-sided thick color steel plates, and the exterior windows are single-layer Low-E glass windows. Specific information on the enclosure structure is shown in Table 1.

[0164] Table 1 Information on Envelope Structure

[0165] Envelope <![CDATA[Heat transfer coefficient (W / (m 2 ·K))]]> Polyurethane insulated panels 0.024 Color steel plate 0.25 Low-E glass windows 3

[0166] The building-integrated photovoltaic (BIPV) system has an installed capacity of 12 kWp. The heating system mainly includes one air-cooled, variable-frequency heat pump, one circulating water pump, one buffer tank, and a radiant floor. Other adjustable appliances include a television, an induction cooker, a dishwasher, a microwave oven, a rice cooker, a coffee maker, a laptop computer, and three adjustable fluorescent lamps. The heat pump unit has a nominal heating capacity of 18 kW and a heating power of 5.77 kW. The trolley has a capacity of 70 kWh, a maximum charge / discharge rate of 0.5C, and a charge / discharge efficiency of 0.9. The battery capacity is 5 kWh. The electricity price uses the residential electricity price for the heating season in a certain province, as detailed in Table 2.

[0167] Table 2 Time-of-use Electricity Price Table

[0168] time Electricity Price Period Electricity price (yuan / kWh) 8：00-20：00 Peak segment 0.8769 20：00-8：00 Valley section 0.6469

[0169] The building energy system is centrally controlled by a cloud server. Based on the Linux system and real-time communication technology, the control hardware is built. The server mainly includes the following modules:

[0170] (1) Data acquisition module.

[0171] The data acquisition module obtains the data required by the model and algorithm through online collection and manual input, and stores it in the database. MySQL is used as its backend database storage.

[0172] Meteorological parameters: Based on weather forecast websites, hourly weather conditions for the next 24 hours, outdoor dry-bulb temperature, solar radiation intensity, relative humidity, air quality, wind speed, wind direction, cloud cover, etc. are obtained through online API technology.

[0173] Main unit information: Obtain the unit model and other information from the unit nameplate.

[0174] Building information: Obtain information such as exterior wall area, exterior wall thickness, door and window area, roof area, roof thickness, and heat exchange terminal type based on building drawings or on-site measurements.

[0175] Time-of-use pricing: Access the State Grid website to obtain the time-of-use pricing for residents in the project area through user queries.

[0176] (2) Each prediction module.

[0177] The various prediction models proposed in this invention were established using the Python language. The photovoltaic power generation prediction model and the heat pump energy consumption prediction model adopted the convolutional neural network algorithm, while the household appliance load prediction model adopted the long short-time memory algorithm.

[0178] Each prediction model was trained and validated. Using 720 sets of operational data from actual buildings from November 15th to 20th, 2024, the particle swarm optimization algorithm was used to identify the parameters of the RC model. The specific parameter identification results are shown in Table 3. Using 2880 sets of operational data from the 2024 heating season, with 70% used for training and 30% for validation, the photovoltaic power generation prediction model, the heat pump energy consumption prediction model, and the remaining household appliance load prediction models were trained and validated respectively, predicting MAPE of 2.1%, 4.3%, and 1.3%, respectively.

[0179] Table 3 RC Model Parameter Identification Results

[0180]

[0181] (3) Optimize the scheduling module.

[0182] Based on the scheduling method proposed in this invention (i.e. Figure 1 We chose Python to develop the optimization strategy, using particle swarm optimization as the first-stage algorithm and MILP as the second-stage algorithm.

[0183] For the first-stage objective function, considering the indoor temperature requirement of ±1℃, the cost weighting factor 'a' is chosen to be 20. During the solution process, considering computational efficiency and accuracy, the population size and maximum number of iterations for the particle swarm optimization algorithm are set to 200 and 1000, respectively.

[0184] The optimization strategy resets the heat pump's water supply temperature. Based on the deviation between the supply temperature and the set temperature, the heat pump automatically adjusts its operating frequency using a PID algorithm.

[0185] The first-stage optimization strategy resets the heat pump's water supply temperature. Based on the deviation between the supply temperature and the set temperature, the heat pump automatically adjusts its operating frequency using a PID algorithm.

[0186] The two-stage optimization strategy adjusts the charging and discharging power of the battery and electric vehicle, as well as the power consumption of other household appliances, with the operating time windows of the other household appliances and the arrival time of the electric vehicle serving as constraints.

[0187] All the modules and databases described above are deployed on a single cloud platform server. In addition, the server is equipped with an external data acquisition API program that automatically obtains future outdoor weather parameters from meteorological websites. Simultaneously, the server can interact with the actual system to collect system operation data and send scheduling commands.

[0188] The indoor thermal environment prediction model and the heat pump energy consumption prediction model are run daily, and the models are updated based on the training results to continuously improve their accuracy. Meteorological data is collected daily or hourly. Time-of-use electricity price data is obtained from the State Grid Corporation of China and updated daily. The photovoltaic power generation prediction model can be run randomly daily, and the model is updated daily for training. The household appliance energy load prediction model runs daily at 22:30, and the model is updated daily for training. The scheduling program is divided into a day-ahead scheduling program and an intraday rolling optimization program, with the day-ahead scheduling program starting at 23:00 daily. The system operation data monitoring and database update frequency is once every 5 minutes.

[0189] Ultimately, this method can reduce the building's heating season operating costs by 15-20%, reduce operating energy consumption by 20-25%, increase photovoltaic absorption rate by 18-25%, and increase heat pump COP by about 5%. Furthermore, the solution speed of this method is about 10 minutes faster than existing algorithms under the same computer configuration.

[0190] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0191] Therefore, this invention employs the aforementioned method for optimizing the scheduling of building energy systems based on efficient utilization of active and passive energy storage. This method integrates photovoltaic power generation, air-source heat pumps, energy storage batteries, reverse-charging trolley piles, adjustable lighting fixtures, and smart home appliances (such as washing machines and water heaters) to achieve efficient energy utilization, cost optimization, and user thermal comfort assurance. By utilizing weather forecasts and building energy system operation data, a four-terminal prediction model is established, encompassing the building's dynamic thermal environment, photovoltaic power generation, appliance load, and heat pump energy consumption. This model enables accurate prediction of both supply and demand for the energy system over the next 24 hours. Combined with a two-stage optimization scheduling method, the first stage focuses on optimizing the heat pump heating water temperature, integrating photovoltaic power generation and time-of-use pricing to ensure thermal comfort. The second stage, based on the remaining photovoltaic power generation capacity, performs refined scheduling of energy storage batteries, reverse-charging piles, adjustable lighting fixtures, and transferable load appliances. This achieves efficient utilization of various flexible resources within the energy system, including building thermal inertia, active energy storage, and adjustable appliances, ultimately maximizing energy efficiency, ensuring thermal comfort, and optimizing economic performance.

[0192] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A building energy system optimal scheduling method based on active and passive energy storage efficient utilization, characterized in that In, comprising the following steps: Step S1, collect building thermal information, heating system operation data, photovoltaic power generation system operation data, household appliance operation data, electric vehicle and energy storage system information; Step S2, establish a four-terminal prediction model, including: building heat transfer dynamic prediction model, heat pump energy consumption prediction model, photovoltaic power generation power prediction model and household appliance load prediction model; Step S3, obtain future 24h weather forecast data and time-of-use electricity price; Step S4, input the data obtained in step S3 into the model in step S2 to obtain the predicted values of indoor temperature, heat load, heat pump energy consumption, photovoltaic power generation power and household appliance load in the next 24 hours; Step S5, perform two-stage optimization scheduling; In step S2, based on the heat resistance and heat capacity analogy principle, a building heat transfer dynamic prediction model is constructed as follows: where C ai represents the heat capacity of indoor air, T ai represents the predicted value of indoor air temperature, t represents time, T w_in represents the temperature of the inner surface of non-transparent envelope, R w_in represents the heat transfer resistance between the inner surface of non-transparent envelope and indoor air, T r_in represents the temperature of the inner surface of non-transparent envelope roof, R r_in represents the heat transfer resistance between the inner surface of non-transparent envelope roof and indoor air, T im represents the temperature of internal heat storage, R im represents the heat transfer resistance between internal heat storage and indoor air, T f represents the temperature of radiant floor, R f represents the heat transfer resistance between radiant floor and indoor air, Q p represents the heat dissipation of personnel / equipment, Q win represents the heat transfer of transparent envelope, C w and C r represent the heat capacity of non-transparent envelope wall and roof, respectively, T w_out and T r_out represent the temperature of the outer surface of non-transparent envelope wall and roof, respectively, T w_in and T r_in represent the temperature of the inner surface of non-transparent envelope wall and roof, respectively, R w and R r represent the heat transfer resistance of non-transparent envelope wall and roof, respectively, R w_out and R r_out represent the heat transfer resistance between the outer surface of non-transparent envelope wall and roof and outdoor air, respectively, R w_in and R r_in represent the heat transfer resistance between the inner surface of non-transparent envelope wall and roof and indoor air, respectively; T ao represents the temperature of outdoor air, Q solar_w and Q solar_r represent the solar radiation heat gain of the surface of non-transparent envelope wall and roof, respectively, C im represents the heat capacity of internal heat storage, Q solar_im represents the solar radiation heat gain through the window and internal heat storage; In step S5, in the first stage, the air source heat pump water supply temperature is dynamically optimized by using bionics algorithm to minimize the system operation cost and indoor temperature deviation, and passive energy storage scheduling is realized by combining building thermal inertia. The objective function of the first stage optimization scheduling is: J(T ws,set,0 ,T ws,set,2 ,...,T ws,set,23 ) = Min(a x C HVAC + ΔT ai ); where J(T ws,set,0 ,T ws,set,2 ,...,T ws,set,23 ) represents a water temperature scheduling target function value, T ws,set represents an optimal air source heat pump water temperature set value, a represents a cost weight factor, and represent minimum and maximum allowable values of the water temperature set, respectively, ΔT ai is a root mean square deviation of the room temperature, C HVAC is a next day heating system operation cost, and the calculation formula is as follows: P HVAC (t) = P ASHP (t) + P PUMP (t); wherein P HVAC (t) represents the system power consumption, P ASHP (t) and the water pump power consumption P PUMP (t) constitute, P HVAC_grid (t) represents the grid power consumption of the heat pump system, P PV (t) is the photovoltaic power generation at time t, y TOU (t) is the time-of-use electricity price at time t; ΔT ai is the root mean square deviation from room temperature, calculated from the next day's indoor hourly temperature T ai (t) and the next day's indoor hourly set temperature T ai,set (t) are calculated. In step S5, in the second stage, based on the remaining amount of photovoltaic power generation that is not consumed in the first stage, the charging and discharging power of the energy storage battery, the charging and discharging power of the electric vehicle and the operating power of the adjustable household appliance are optimized by using mixed integer linear programming algorithm. The objective function of the second stage optimization scheduling is: J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j1,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 )=Min(C sys ); wherein, J(P c,0 ,P dc,0 ,P ev_c,0 ,P ev_dc,0 ,P ad,z,0 ,P sh,j,0 ...,P c,23 ,P dc,23 ,P ev_c,23 ,P ev_dc,23 ,P ad,z,23 ,P sh,j,23 ) represents a two-stage building energy system scheduling objective function, P c ,P dc represents the optimal battery charging and discharging power, P ev_c ,P ev_dc represents the optimal electric vehicle charging and discharging power, P ad,z represents the power consumption of the zth adjustable household appliance, P sh,j represents the power consumption of the jth transfer load household appliance; C sys The next day building energy system operation cost is solved by the following formula: where P grid (t) represents the amount of electricity that needs to be purchased from the grid after the optimized scheduling, p(t) represents the time-of-use electricity price at time t, and the solving process establishes a power balance model as follows: P grid (t)+P dc (t)+P ev_dc (t)≥P HVAC (t); P grid (t)+P dc (t)+P ev_dc (t)+P pv_unused (t)≥P app (t)+P app_fle (t)+P c (t)+P ev_c (t); where P pv_unused (t) represents the photovoltaic power not absorbed in the first stage, P c (t), P ev_c (t) and P dc (t), P ev_dc (t) represents the charging and discharging power of the battery and the electric vehicle, P app (t) and P app_fle (t) represents the basic load and flexible load of the remaining home appliances.

2. The building energy system optimal scheduling method based on active and passive energy storage efficient utilization according to claim 1, characterized in that, In step S1, the building thermal information includes the size, material and heat transfer coefficient of the building envelope; the heating system operation data includes the heating system water temperature, outdoor air temperature, solar radiation intensity, indoor air temperature and energy consumption; the photovoltaic power generation system operation data includes outdoor air temperature, solar radiation intensity, outdoor air humidity, cloud cover and photovoltaic power generation power; the household appliance operation data includes the type of household appliance, the use time range of household appliance, the adjustable power and the total electrical load of household appliance; the electric vehicle and energy storage system information includes the battery capacity, the charging and discharging depth, the charging and discharging power range, the charging and discharging efficiency and the initial state of charge.

3. The method of claim 1, wherein, In step S2, a neural network is used to establish a heat pump energy consumption prediction model, the input parameters of which are outdoor temperature, water supply temperature and return water temperature, and the output parameter is the unit energy consumption; A neural network is used to establish a photovoltaic power generation power prediction model, the input parameters of which are solar radiation intensity, outdoor temperature, humidity and cloud cover, and the output parameter is photovoltaic power generation power; A neural network is used to establish a household appliance load prediction model, the input parameters of which are outdoor temperature and humidity and historical electrical load.

4. The building energy system optimal scheduling method based on active and passive energy storage efficient utilization according to claim 1, characterized in that, P app_fle The solution of (t) is mainly divided into two kinds: power adjustable equipment flexibility and load transferable equipment flexibility. P app_fle (t) = P app_fle,ad (t) + P app_fle,sh (t); t work,s ≥t window,s t work,e ≤t window,e l work ≤l window ; wherein N z (t) represents the number of power adjustable home appliances of the z-th type that can be adjusted at time t, M j (t) represents the number of load transferable home appliances of the j-th type that can be adjusted at time t, P ad,z (t) represents the power of the z-th power adjustable home appliance that can be reduced at time t, P sh,j (t) represents the running power of the j-th load transferable home appliance, U j (t) represents the flexibility state of the j-th load transferable home appliance at time t, n represents the number of power adjustable home appliances that can be adjusted at time t, m represents the number of load transferable home appliances that can be adjusted at time t, P app_fle,ad (t) represents the flexible load that can be provided by the adjustable home appliance at time t, P app_fle,sh (t) represents the flexible load that can be provided by the transferable home appliance at time t, l work,shift represents the time period that can be transferred within the working window, t work,s represents the start time of the equipment working, t window,s represents the start time of the time window, t work,e represents the end time of the equipment working, t window,e represents the end time of the time window, l work represents the working time length, l window,s represents the time length of the time window; The flexible calculation formula is as follows: Q app_fle (t) = Q app_fle,ad (t) + Q app_fle,sh (t); wherein n z represents the heat dissipation coefficient of the zth power-adjustable home appliance, represents the cooling load coefficient of the heat dissipation of the zth power-adjustable home appliance, C sh,j represents the operating power of the jth load-transferable home appliance, n1 represents the ballast consumption power coefficient of the fluorescent lamp, n2 represents the lamp cover heat insulation coefficient of the fluorescent lamp, n3 represents the motor utilization coefficient, n4 represents the motor load coefficient, n5 represents the heat dissipation coefficient of the electric heating device or electronic device, η represents the motor efficiency, Q app_fle (t) represents the total flexible load generated by the heat dissipation of the remaining home appliances at time t, Q app_fle,ad (t) represents the total flexible load that can be provided by the heat dissipation of the power-adjustable home appliances at time t, Q app_fle,sh (t) represents the total flexible load that can be provided by the heat dissipation of the power-transferable home appliances at time t, COP represents the working performance coefficient of the home appliance.

5. The method of claim 4, wherein, The charge-discharge power P of the battery c (t) and P dc(t) subject to the following battery scheduling process constraints: where SOC (t) represents the current energy state of the battery, SOC(t0) represents the energy state of the battery at the initial time t0, η c and η dc represent the charging and discharging efficiency of the battery, Δt represents the time interval, E Max is the rated capacity of the battery; The battery is constrained by the maximum and minimum capacity allowed during scheduling, specifically: SOC min ≤ SOC(t) ≤ SOC max ; where SOCmaxand SOCminrepresent the maximum capacity and the minimum capacity of the battery, respectively. max and SOCminrepresent the maximum capacity and the minimum capacity of the battery, respectively. min where SOCmaxand SOCminrepresent the maximum capacity and the minimum capacity of The battery is constrained by the maximum and minimum power allowed during charging and discharging, specifically: wherein, and Pmaxdenotes the maximum charge and discharge power of the battery; Since the battery cannot perform simultaneous charging and discharging, a binary variable is introduced to constrain the mutual exclusion of charging and discharging states: x c +x dc ≤1; where x c and x dc denote the state of charge 0 and 1 of the battery; Charging and discharging cycle constraint: Where T represents the time range covered by the optimization scheduling.

Images