A public building flexible load virtual energy storage joint scheduling method and system for distributed photovoltaic on-site consumption

Abstract

The application discloses a kind of public building flexible load virtual energy storage combined dispatching method and system for distributed photovoltaic in-situ consumption, the method by establishing single air conditioner dynamic virtual energy storage model, the uncertainty factor of air conditioner cluster is aggregated by monte carlo simulation to realize air conditioner cluster dynamic virtual energy storage model;According to the probability density function of electric vehicle arrival public building time, electric vehicle leaves public building time and electric vehicle arrival public building initial state of charge, generate electric vehicle cluster scene based on monte carlo, construct electric vehicle cluster dynamic virtual energy storage model;Air conditioner cluster and electric vehicle cluster dynamic virtual energy storage index are constructed, air conditioner virtual energy storage and electric vehicle virtual energy storage are jointly optimized, significantly reduce the decision variable of joint optimization process, realize the overall scheduling of air conditioner and electric vehicle, improve building comprehensive energy efficiency and realize green low-carbon operation.

Description

Technical Field

[0001] This invention relates to the field of demand-side optimization control of integrated energy systems, and in particular to a method and system for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power. Background Technology

[0002] With the increasing integration of photovoltaic (PV) energy into buildings, optimizing the control of building flexible loads to fully utilize PV energy has become a hot research topic. PV power generation is characterized by output uncertainty and high volatility, thus requiring buildings to possess sufficient response potential for rapid adjustment. Building virtual energy storage (BVA) technology is an accurate method for quantifying building regulation potential, effectively helping buildings to locally utilize renewable energy, reducing investment in building energy storage configurations, and promoting energy conservation and emission reduction. However, current research on BVA mainly focuses on single virtual energy storage devices, such as optimizing virtual energy storage for a single building's air conditioning system or electric vehicle, without considering the synergistic relationships between various virtual energy storage systems or the enormous potential and application prospects of BVA in promoting local PV utilization. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide a method for joint scheduling of virtual energy storage for flexible loads in public buildings based on air conditioning clusters and electric vehicle clusters for local consumption of distributed photovoltaic power; another purpose of this invention is to provide a system for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power.

[0004] Technical solution: The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in this invention, includes the following steps:

[0005] (1) Considering temperature changes, a dynamic virtual energy storage model for a single air conditioner is constructed based on the dynamic virtual energy storage index of the air conditioner; the dynamic virtual energy storage index of the air conditioner includes the virtual charging and discharging power of the air conditioner, the virtual capacity of the air conditioner and the virtual state of charge of the air conditioner.

[0006] (2) Based on the Monte Carlo distribution of the start-up time, continuous operation time and optimal set temperature of the air conditioners in each room of the public building, generate a Monte Carlo-based air conditioner cluster scenario and construct a dynamic virtual energy storage model for the air conditioner cluster.

[0007] (3) Based on the probability density function of the arrival time of electric vehicles in public buildings, the departure time of electric vehicles from public buildings, and the initial state of charge of electric vehicles upon arrival in public buildings, generate a Monte Carlo-based electric vehicle cluster scenario and construct a dynamic virtual energy storage model for the electric vehicle cluster.

[0008] (4) According to the dynamic virtual energy storage index of the air conditioner, the virtual charging and discharging power of the electric vehicle cluster, a public building coordinated operation optimization objective function is constructed, and a dynamic multi-objective particle swarm algorithm is used to dynamically solve the weight of the objective function to obtain a flexible load joint scheduling strategy for distributed photovoltaic on-site consumption.

[0009] Further, in step (1), the air conditioner virtual charging and discharging power satisfies the following conditions:

[0010] ;

[0011] In the formula, P ac is the actual operating power of the air conditioner, in kW; P ac,base is the basic power required to maintain the indoor temperature at the set temperature T set , in kW; P ac,ch / P ac,dis respectively represent the air conditioner virtual charging and discharging power, in kW;

[0012] The air conditioner virtual capacity is

[0013] ;

[0014] ;

[0015] In the formula, is the air conditioner virtual capacity in the Lth optimization period, in kWh; t L is the air conditioner off time, in h; is the time for the indoor temperature to rise from the minimum indoor temperature T min to the maximum indoor temperature T max in the Lth optimization period when the air conditioner is off, in h; C is the equivalent heat capacity coefficient of the building, in KJ / ℃; λ o is the convective heat transfer coefficient of the outdoor environment and the external wall, in W / (m2·℃); λ i is the convective heat transfer coefficient of the indoor environment and the external wall; A0 is the total area of the external heat storage material, in m2; h i is the heat exchange coefficient between the ith heat storage material and the indoor environment, in W / (m2·℃); T0 is the outdoor temperature, in ℃; T max is the maximum indoor temperature, in ℃; T min is the minimum indoor temperature, in ℃;

[0016] The air conditioner virtual state of charge SOC ac (L) is

[0017] ;

[0018] ;

[0019] ;

[0020] E(L) is the virtual energy storage of air conditioner in the Lth optimization period, unit: kWh; The value range of is 0~1, T i is the current indoor temperature, unit: ℃; is the Lth optimization period, the time required for the indoor temperature to rise from the minimum indoor temperature T min to the current indoor temperature T i when the air conditioner is off, unit: h.

[0021] Further, in step (2), the Monte Carlo distribution of the air conditioner start time of each room of the public building is as follows:

[0022] Time period A (09:00-18:00):

[0023] ;

[0024] Time period B (18:00-09:00 of the next day):

[0025] ;

[0026] In the formula, is the start time of the air conditioner of each room of the public building, unit: h; is the air conditioner start time probability density function of each room of the public building in time period A; is the air conditioner start time probability density function of each room of the public building in time period B;

[0027] The Monte Carlo distribution of the continuous running time of the air conditioner of each room of the public building is as follows:

[0028] Time period A (09:00-18:00):

[0029] ;

[0030] Time period B (18:00-09:00 of the next day):

[0031] ;

[0032] In the formula, is the continuous running time of the air conditioner of each room of the public building, unit: h; is the air conditioner continuous running time probability density function of each room of the public building in time period A; is the air conditioner continuous running time probability density function of each room of the public building in time period B;

[0033] The Monte Carlo distribution of the optimal set temperature of each room of the public building is as follows:

[0034] ;

[0035] In the formula, is the optimal set temperature of air conditioning of each room of the public building, in ℃, ; is the probability density function of the optimal set temperature of each room of the public building.

[0036] Further, in step (2), the total virtual charging and discharging power of the air conditioner cluster in the dynamic virtual energy storage model of the air conditioner cluster is

[0037] ;

[0038] In the formula, is the total virtual charging and discharging power of the air conditioner cluster, in kW; is the basic power required for the i th air conditioner to maintain the indoor temperature as the set temperature T set , in the L th optimization period, in kW; is the charging power of the i th air conditioner in the L th optimization period, in kW; is the discharging power of the i th air conditioner in the L th optimization period, in kW; and N is the total number of air conditioners. L i

[0039] The total virtual energy storage capacity of the air conditioner cluster is

[0040] ;

[0041] In the formula, is the total virtual energy storage capacity of the air conditioner cluster, in kWh; is the virtual energy storage capacity of the i th air conditioner in the L th optimization period, in kWh;

[0042] The total virtual energy storage capacity of the air conditioner cluster is

[0043] ;

[0044] In the formula, is the total virtual energy storage capacity of the air conditioner cluster, in kWh; is the virtual energy storage capacity of the i th air conditioner in the L th optimization period, in kWh;

[0045] The total virtual state of charge is ;

[0046] In the formula,​​ The value range of the Lth optimization period is 0-1.

[0047] Further, in step (3), the time for the electric vehicle to arrive at the public building satisfies the following condition:

[0048] ;

[0049] wherein, is the probability density function of the time for the electric vehicle to arrive at the public building; is the mean of the time for the electric vehicle to arrive at the public building, is the variance of the time for the electric vehicle to arrive at the public building, is the time for the electric vehicle to arrive at the public building, in h;

[0050] The time for the electric vehicle to leave the public building satisfies the following condition:

[0051] ;

[0052] wherein, is the probability density function of the time for the electric vehicle to leave the public building; μ r is the mean of the time for the electric vehicle to leave the public building, σ r is the variance of the time for the electric vehicle to leave the public building, is the time for the electric vehicle to leave the public building, in h;

[0053] The initial state of charge of the electric vehicle arriving at the public building satisfies the following condition:

[0054] ;

[0055] wherein: is the probability density function of the initial state of charge of the electric vehicle arriving at the public building; μ B is the mean of the initial state of charge of the electric vehicle arriving at the public building, σ B is the variance of the initial state of charge of the electric vehicle arriving at the public building, is the initial state of charge of the electric vehicle arriving at the public building, with a value range of 0-1.

[0056] Further, in step (3), the virtual charging and discharging power of the electric vehicle cluster of the electric vehicle cluster dynamic virtual energy storage model satisfies the following condition:

[0057] ;

[0058] ;

[0059] ;

[0060] wherein, and are the virtual charging and discharging power of the Lth optimization period of the electric vehicle cluster, respectively, in kW; is the total power of the Lth optimization period of the electric vehicle, in kW; is the charging power of the jth electric vehicle in the Lth optimization period, in kW; is the discharging power of the jth electric vehicle in the Lth optimization period, in kW; is the charging and discharging state of the jth electric vehicle in the Lth optimization period, 1 for charging state and 0 for discharging state; N EV is the total number of electric vehicles;

[0061] The virtual energy storage charging and discharging capacity of the electric vehicle cluster satisfies the following conditions:

[0062] ;

[0063] ;

[0064] wherein, and are the virtual energy storage charging and discharging capacity of the Lth optimization period, respectively, in kWh; N EV is the number of electric vehicles participating in optimization; is the state of charge of the jth electric vehicle in the Lth optimization period; is the maximum expected state of charge of the user of the electric vehicle; is the minimum expected state of charge of the user of the electric vehicle; E is the rated battery capacity of the electric vehicle;

[0065] is the total virtual capacity of the electric vehicles participating in operation optimization in the Lth optimization period is

[0066] ;

[0067] is the total virtual state of charge of the electric vehicle cluster

[0068] ;

[0069] wherein, is the total SOC of the virtual energy storage of the electric vehicle in the Lth optimization period, with a value range of 0-1.

[0070] Further, in step (4), according to the dynamic virtual energy storage index of the air conditioner and the dynamic virtual energy storage index of the electric vehicle cluster, a public building coordinated operation optimization objective function is constructed as follows:

[0071] ;

[0072] a is the air conditioning operation and maintenance price, unit: yuan / kWh; are the charging / discharging price of the electric vehicle in the Lth optimization period, unit: yuan / kWh; is the length of an optimization period, unit: h; p(L) is the grid electricity price, unit: yuan / kWh; c(L) is the carbon emission penalty cost in the Lth optimization period, unit: yuan / kg; is the dynamic carbon emission factor in the Lth optimization period, unit: kg / kWh; is the comfort deviation penalty cost, unit: yuan; is the photovoltaic predicted output in the Lth optimization period, unit: kW; is the actual photovoltaic usage in the Lth optimization period, unit: kW; is the weight coefficient of the two objectives; is the building external power purchase in the Lth optimization period, unit: kW; is the PMV value corresponding to the indoor temperature in the Lth optimization period; is the PMV value corresponding to the most suitable indoor temperature; T is the number of optimization periods.

[0073] Further, in step (4), the dynamic multi-objective particle swarm algorithm is used to dynamically solve the objective function weight, and in N times of solving, the weight coefficient solving formula is

[0074]

[0075] The updating formula of the particle motion speed is

[0076]

[0077] In the formula, v is the speed of the particle in the Nth iteration calculation; v is the speed of the particle in the (N+1)th iteration calculation; r1 and r2 are random numbers in the interval [0, 1] and independent of each other; x is the current optimal position of the particle; x is the global optimal position; inertia constant is taken in the interval [0, 1] linearly decreasing, is the position of the i th particle in the Nth iteration; is the cognitive learning factor for adjusting the individual experience weight; is the social learning factor for adjusting the social experience weight. ​​​​​​​​​

[0078] Further, the updating formula of the particle position is

[0079] ;

[0080] In the formula, is the position of the i th particle at the N th iteration.

[0081] The public building flexible load virtual energy storage joint scheduling system for distributed photovoltaic on-site consumption provided by the application comprises a single air conditioner dynamic virtual energy storage model construction module, which is used for considering temperature changes, constructing a single air conditioner dynamic virtual energy storage model according to dynamic virtual energy storage indexes of the air conditioner, wherein the dynamic virtual energy storage indexes of the air conditioner include air conditioner virtual charging and discharging power, air conditioner virtual capacity and air conditioner virtual state of charge; an aggregated air conditioner virtual energy storage model construction module, which is used for generating an air conditioner cluster scene based on Monte Carlo according to the time of air conditioner starting, the continuous operation time of the air conditioner and the Monte Carlo distribution of the most suitable set temperature of each room of the public building, and constructing an air conditioner cluster dynamic virtual energy storage model; an electric vehicle cluster aggregated dynamic virtual energy storage model construction module, which is used for generating an electric vehicle cluster scene based on Monte Carlo according to the time of electric vehicles arriving at the public building, the time of electric vehicles leaving the public building and the probability density function of the initial state of charge of the electric vehicles arriving at the public building, and constructing an electric vehicle cluster dynamic virtual energy storage model; and a strategy solving module, which is used for constructing a public building coordinated operation optimization objective function according to the dynamic virtual energy storage indexes of the air conditioner, the virtual charging and virtual discharging power of the electric vehicle cluster, dynamically solving the weights of the objective function by using a dynamic multi-objective particle swarm algorithm, and obtaining a flexible load joint scheduling strategy for distributed photovoltaic on-site consumption.

[0082] Advantages: Compared with the prior art, the application has the following advantages: 1. The application aggregates the air conditioner cluster dynamic virtual energy storage model by simulating the air conditioner cluster through Monte Carlo according to the single air conditioner dynamic virtual energy storage model; 2. The application constructs an electric vehicle cluster dynamic virtual energy storage model by generating an electric vehicle cluster scene based on Monte Carlo according to the time of electric vehicles arriving at the public building, the time of electric vehicles leaving the public building and the probability density function of the initial state of charge of the electric vehicles arriving at the public building; 3. The application constructs dynamic virtual energy storage indexes of the air conditioner cluster and the electric vehicle cluster, jointly optimizes the air conditioner virtual energy storage and the electric vehicle virtual energy storage, significantly reduces the decision variables in the joint optimization process, realizes the overall scheduling of the air conditioner and the electric vehicle, and improves the building comprehensive energy efficiency and realizes green and low-carbon operation. BRIEF DESCRIPTION OF DRAWINGS

[0083] Figure 1 It is a structural schematic diagram of the public building;

[0084] Figure 2 The schematic diagram of the virtual energy storage equivalent structure for the building;

[0085] Figure 3 The column chart of the virtual charging and discharging power of the air conditioner;

[0086] Figure 4 The column chart of the virtual energy storage capacity and SOC of the air conditioner;

[0087] Figure 5 The column chart of the virtual charging and discharging power of the electric vehicle;

[0088] Figure 6 The column chart of the virtual capacity and SOC of the electric vehicle;

[0089] Figure 7 The line chart of the temperature setting and indoor and outdoor temperature of the air conditioner cluster after the joint optimization strategy;

[0090] Figure 8 The column chart of the photovoltaic usage and photovoltaic utilization rate. DETAILED DESCRIPTION

[0091] The application will be further described below with reference to the drawings.

[0092] The public building flexible load virtual energy storage joint scheduling method for distributed photovoltaic on-site consumption, comprising the following steps: considering temperature change, constructing a single air conditioner dynamic virtual energy storage model according to the dynamic virtual energy storage index of the air conditioner; wherein the dynamic virtual energy storage index of the air conditioner includes air conditioner virtual charging and discharging power, air conditioner virtual capacity and air conditioner virtual state of charge.

[0093] The air conditioner virtual energy storage is also modeled from the perspective of electrical energy storage. Analogous to the parameters of the traditional energy storage model, the air conditioner virtual energy storage has three indexes: virtual charging and discharging power, virtual capacity and virtual state of charge.

[0094] When the outdoor temperature rises, the indoor temperature T i To maintain the air conditioner setting temperature T set , the required power also increases. Conversely, when the outdoor temperature decreases, the required power also decreases. According to this temperature change characteristic, the operating power of the air conditioner is defined as P ac , and the charging and discharging process of the air conditioner virtual energy storage can be described as: when P ac > , the air conditioner virtual energy storage is charged; when P ac < , the air conditioner virtual energy storage is discharged. The actual power of the air conditioner is represented by the air conditioner charging and discharging power and the air conditioner reference power:

[0095] ;

[0096] In the formula, P ac This refers to the actual operating power of the air conditioner, measured in kW; P ac,base It maintains the indoor temperature at the set temperature T. set The required basic power, in kW; P ac,ch / P ac,dis These represent the virtual charging and discharging power of the air conditioner, in kW.

[0097] Battery capacity is a parameter reflecting the amount of electricity a battery can store, usually expressed as the integral of the battery's power over a period of time. The virtual energy storage model for air conditioners uses a similar method; the virtual capacity is calculated as the integral of the room's heat dissipation power over that period. The virtual capacity of the air conditioner in the Lth optimization period is shown in the formula: ;

[0098] In the formula, The virtual power capacity of the air conditioner is expressed in kWh; L represents the Lth optimization period; t L This is the time the air conditioner is turned off, measured in hours (h). When the air conditioner is turned off during the Lth optimization period, the indoor temperature drops from the lowest indoor temperature T. min Rise to the highest indoor temperature T max The time, in hours;

[0099] ;

[0100] In the formula, C is the equivalent heat capacity coefficient of the building, with units of KJ / ℃; λ o The convective heat transfer coefficient between the outdoor environment and the exterior wall is expressed in W / (m²). 2 ·℃); λ i A0 represents the convective heat transfer coefficient between the indoor environment and the exterior walls; A0 represents the total area of ​​the external heat storage material, in m2. 2 h i The coefficient for heat exchange between the i-th heat source and the room is expressed in W / (m²). 2 ·℃); T0 is the outdoor temperature, in ℃; T max The highest indoor temperature, expressed in °C; T min T0 represents the lowest indoor temperature, measured in °C; T0 represents the outdoor temperature.

[0101] The state of charge (SBC) of a battery reflects the amount of remaining battery power, typically described as the ratio of remaining battery power to battery capacity. In the virtual energy storage model, the SBC is defined as the ratio of stored electrical energy to the virtual capacitance of the air conditioner, reflecting the margin of the air conditioner's heating capacity. The virtual SBC for the Lth optimization period is: ;

[0102] wherein, E(L) is the virtual energy storage of air conditioner in the Lth optimization period, unit: kWh; The value range of is 0~1.

[0103] The virtual energy storage of air conditioner E(L) is defined as the time when the Lth optimization period is closed, the indoor temperature rises from the minimum T min to the current indoor temperature T i required. The integral of this time is:

[0104] ;

[0105] ;

[0106] wherein, T i is the current indoor temperature, unit: ℃; is the Lth optimization period, the time required for the indoor temperature to rise from the minimum indoor temperature T min to the current indoor temperature T i , unit: h.

[0107] (1) According to the Monte Carlo distribution of the air conditioner starting time, continuous running time and the most suitable set temperature of each room in the public building, a Monte Carlo-based air conditioner cluster scenario is generated to construct the dynamic virtual energy storage model of the air conditioner cluster.

[0108] Monte Carlo distribution of the starting time of each room in the public building: when constructing the multi-operation scenario of the air conditioner cluster, it is necessary to analyze the starting time of the air conditioner of each room in the public building first. Through the analysis of the historical operation data of the building, it is found that the air conditioner starting time presents different distribution rules in different time periods. In order to more accurately simulate the actual situation, one day is divided into two main time periods: time period A (09:00-18:00): this is the main working time of the public building, and the air conditioner starts frequently. According to the historical data statistics, the probability density function of the air conditioner starting time can be expressed as:

[0109] ;

[0110] Time period B (18:00-09:00 the next day): this is the non-main working time, and the air conditioner starts relatively less. The probability density function of the air conditioner starting time can be expressed as:

[0111] ;

[0112] wherein, is the starting time of the air conditioner of each room in the public building, unit: h; Probability density function of the air-conditioning start-up time for each room of the public building in time period A; Probability density function of the air-conditioning start-up time for each room of the public building in time period B.

[0113] Monte Carlo distribution of the continuous running time for each room of the public building: In addition to the start-up time, the continuous running time of the air-conditioning is also an important factor affecting the energy consumption. The continuous running time reflects the length of time that the air-conditioning works continuously after starting, and its distribution is also different in different time periods.

[0114] Time period A (09:00-18:00): In the main working time period, the air-conditioning usually needs to run continuously for a long time to maintain indoor comfort. The probability density function of the continuous running time can be expressed as:

[0115]

[0116] Time period B (18:00-09:00 of the next day): In the non-main working time period, the continuous running time of the air-conditioning is relatively short. The probability density function of the continuous running time can be expressed as:

[0117]

[0118] wherein, is the continuous running time of the air-conditioning for each room of the public building, in h; is the probability density function of the air-conditioning continuous running time for each room of the public building in time period A; is the probability density function of the air-conditioning continuous running time for each room of the public building in time period B.

[0119] Monte Carlo distribution of the optimal set temperature for each room: The optimal set temperature of different rooms is different according to their functions and use requirements. In order to simulate this diversity, the temperature Monte Carlo distribution is set, which is subject to a uniform distribution:

[0120] wherein, is the optimal set temperature of the air-conditioning for each room of the public building, in ℃; is the probability density function of the optimal set temperature for each room of the public building.

[0121] This distribution shows that the optimal set temperature of each room is between 23 ℃ and 29 ℃, with an interval of 0.5 ℃, and the probability of each temperature value is 1 / 13, reflecting the diversity and uniformity of the set temperature.

[0122] ​​​​After obtaining the Monte Carlo distribution of the starting time, continuous running time and the most suitable set temperature of the air conditioning cluster, the independent running scenario sample set of each air conditioner in the air conditioning cluster is generated through 500 times of sampling , each sample contains parameters , N is the number of air conditioners in the air conditioning cluster.

[0123] Firstly, the starting time, continuous running time and the most suitable set temperature of each air conditioner in the building are obtained through Monte Carlo simulation, and then aggregation is performed to construct an aggregated air conditioner virtual energy storage model.

[0124] Total virtual charge and discharge power: the total virtual charge and discharge power refers to the sum of the powers of all air conditioners in the cooling and heating modes, reflecting the overall power balance of the air conditioning system:

[0125] ;

[0126] In the formula, is the basic power required for the i-th air conditioner to maintain the indoor temperature as the set temperature T set in the L-th optimization period, with the unit of kW; is the charging power of the i-th air conditioner in the L-th optimization period, with the unit of kW; is the discharging power of the i-th air conditioner in the L-th optimization period, with the unit of kW.

[0127] Total virtual energy storage capacity: the total virtual capacity is the sum of the current virtual capacities of all air conditioners, reflecting the current energy storage state of the air conditioning system:

[0128] ;

[0129] In the formula: is the total virtual energy storage capacity of the air conditioning cluster, with the unit of kWh; is the virtual energy storage capacity of the i-th air conditioner in the L-th optimization period, with the unit of kWh.

[0130] The total virtual energy storage capacity is the maximum virtual energy storage capacity of all air conditioners.

[0131] ;

[0132] In the formula: is the total virtual energy storage capacity of the air conditioning cluster, with the unit of kWh; is the virtual energy storage capacity of the i-th air conditioner in the L-th optimization period, with the unit of kWh.

[0133] The total virtual state of charge is an important indicator for measuring the energy storage state of the air conditioning system, reflecting the ratio of the energy storage capacity of the air conditioning system to the maximum energy storage capacity:

[0134] ;

[0135] In the formula: SOCL is the state of charge of the air conditioning cluster dynamic virtual energy storage in the Lth optimization period, with a value range of 0-1.

[0136] (2) According to the probability density function of the electric vehicle arrival time at the public building, the electric vehicle departure time from the public building and the electric vehicle initial state of charge when arriving at the public building, a Monte Carlo-based electric vehicle cluster scenario is generated, and an electric vehicle cluster dynamic virtual energy storage model is constructed.

[0137] The electric vehicle arrival time at the public building satisfies the following condition:

[0138] ;

[0139] In the formula: is the probability density function of the electric vehicle arrival time at the public building; μ a = 8.86, σ a = 3.19, is the electric vehicle arrival time at the public building, with a unit of h.

[0140] The electric vehicle departure time from the public building satisfies the following condition:

[0141] ;

[0142] In the formula: is the probability density function of the electric vehicle departure time from the public building; μ r = 17.20, σ r = 3.39, is the electric vehicle departure time from the public building, with a unit of h.

[0143] The electric vehicle initial state of charge when arriving at the public building satisfies the following condition:

[0144] ;

[0145] In the formula: is the probability density function of the electric vehicle initial state of charge when arriving at the public building; μ B = 0.5137, σ B = 0.1772, is the electric vehicle initial state of charge when arriving at the public building, with a value range of 0-1.

[0146] After obtaining the Monte Carlo distribution of the electric vehicle cluster arrival time, departure time and initial state of charge when arriving at the public building, through 500 times of sampling, an independent scenario sample set of each electric vehicle in the electric vehicle cluster is generated , each sample contains parameters , is the number of electric vehicles in the electric vehicle cluster.

[0147] The virtual charging and discharging power of the electric vehicle cluster dynamic virtual energy storage model.

[0148] The virtual charging and discharging power of the electric vehicle can be expressed as the sum of the overall charging and discharging power of the electric vehicle:

[0149] ;

[0150] ;

[0151] ;

[0152] In the formula, and respectively represent the virtual charging and virtual discharging power of the electric vehicle cluster in the L period, with the unit of kW; is the total power of the electric vehicle at time t, with the unit of kW; is the charging power of the jth electric vehicle in the L period, with the unit of kW; is the discharging power of the jth electric vehicle in the L period, with the unit of kW; is the charging and discharging state of the jth electric vehicle in the L period, 1 for charging state and 0 for discharging state.

[0153] According to the SOC value of each vehicle at this moment, the single vehicle capacity at this moment can be calculated, and the minimum capacity limit is subtracted, aggregated, and the virtual capacity of the electric vehicle can be calculated.

[0154] ;

[0155] ;

[0156] In the formula, and are the charging and discharging capacity of the available energy storage in the Lth optimization period, with the unit of kWh; N EV is the number of electric vehicles participating in optimization; is the state of charge of the jth electric vehicle in the Lth optimization period; is the maximum expected state of charge of the user of the electric vehicle; is the minimum expected state of charge of the user of the electric vehicle; E is the rated battery capacity of the electric vehicle.

[0157] The total capacity of the electric vehicles that can participate in operation optimization at time t is:

[0158] ;

[0159] Similar to the calculation method of actual energy storage, the virtual state of charge of the electric vehicle is defined as the ratio of the total stored energy to the total virtual capacity, which is used to reflect the adjustable margin of the electric vehicle:

[0160] ;

[0161] In the formula: is the total SOC of the virtual energy storage of the electric vehicle in the Lth optimization period, with a value range of 0-1.

[0162] (3) According to the dynamic virtual energy storage index of the air conditioner, the virtual charging and discharging power of the electric vehicle cluster, the coordinated operation optimization objective function of the public building is constructed, and the dynamic multi-objective particle swarm algorithm is used to dynamically solve the weight of the objective function, to obtain the flexible load joint scheduling strategy for distributed photovoltaic on-site consumption.

[0163] In the actual daily operation of public buildings, photovoltaic has great uncertainty. The actual power generation of photovoltaic is affected by many factors, such as the change of weather, outdoor temperature, seasonal change, etc., which will cause the fluctuation of photovoltaic power generation. The coordinated operation optimization model of air conditioner virtual energy storage and electric vehicle virtual energy storage is introduced, which can store energy when photovoltaic is sufficient and release energy when photovoltaic is insufficient, realize the time and space transfer of energy, and effectively improve the green energy consumption capacity of the building.

[0164] The coordinated operation optimization model of the air conditioner cluster and the electric vehicle cluster should consider both the green and low-carbon operation of the public building and the maximum consumption of photovoltaic. On the one hand, the operation and maintenance cost of the air conditioner cluster, the charging and discharging optimization cost of the electric vehicle, the cost of purchased electricity, the carbon emission cost and the penalty cost of deviating from the optimal temperature are considered; on the other hand, the maximum consumption of photovoltaic is considered. The two aspects are considered comprehensively and weighted. The overall coordinated operation optimization model is as follows:

[0165] ;

[0166] In the formula, a is the operation and maintenance price of the air conditioner cluster, with a unit of yuan / kWh; / are the charging / discharging prices of the electric vehicle in the Lth optimization period, with a unit of yuan / kWh; is the length of an optimization period, with a unit of h; p(L) is the electricity price of the power grid, with a unit of yuan / kWh; c(L) is the carbon emission penalty cost in the Lth optimization period, with a unit of yuan / kg; is the dynamic carbon emission factor in the Lth optimization period, with a unit of kg / kWh; Comfort deviation penalty cost, unit: yuan; Photovoltaic predicted output for the Lth optimization period, unit: kW; Photovoltaic actual usage for the Lth optimization period, unit: kW; Weight coefficient of two objectives; Building external purchased electricity for the Lth optimization period, unit: kW; PMV value corresponding to indoor temperature for the Lth optimization period; PMV value corresponding to optimal indoor temperature; T is the number of optimization periods.

[0167] The joint scheduling adopts a dynamic multi-objective particle swarm algorithm to dynamically solve the weight of the objective function. In N times of solving, the solving formula of the weight coefficient is:

[0168] ;

[0169] The updating formula of the particle motion speed is:

[0170] ;

[0171] In the formula: is the speed of the particle in the Nth iteration calculation; is the speed of the particle in the N+1th iteration calculation; and are random numbers in the interval [0, 1] that are independent of each other; is the current optimal position of the particle; is the global optimal position; inertia constant takes a value in the interval [0, 1] linearly decreasing, so that the search space is constantly reduced. is the position of the ith particle in the Nth iteration; is a cognitive learning factor for adjusting the weight of individual experience; is a social learning factor for adjusting the weight of social experience.

[0172] The updating formula of the particle position is: ; In the formula: is the position of the ith particle in the Nth iteration.

[0173] Through the above iteration formula, the weight of the objective can be changed constantly in the calculation process, so as to guide the particle swarm to move on the Pareto optimal frontier and obtain a Pareto optimal solution set. In these solutions, one is selected as the best compromise solution to meet the balance requirements between multiple objectives.

[0174] The present invention discloses a joint scheduling system for virtual energy storage of flexible loads in public buildings for distributed photovoltaic local consumption. This system includes a module for constructing dynamic virtual energy storage models for individual air conditioners, which considers temperature changes and constructs a dynamic virtual energy storage model for each air conditioner based on its dynamic virtual energy storage indicators. These indicators include virtual charging and discharging power, virtual capacity, and virtual state of charge. The system also includes a module for constructing aggregated virtual energy storage models for air conditioners, which generates a Monte Carlo-based air conditioner cluster scenario based on the Monte Carlo distribution of the start-up time, continuous operation time, and optimal set temperature of the air conditioners in each room of the public building, and constructs a dynamic virtual energy storage model for the air conditioner cluster. The system includes: a dynamic virtual energy storage model construction module for electric vehicle cluster aggregation, used to generate a Monte Carlo-based electric vehicle cluster scenario based on the probability density function of electric vehicle arrival time, electric vehicle departure time, and initial state of charge of electric vehicles upon arrival at the public building; and a strategy solution module, used to construct an optimization objective function for coordinated operation of public buildings based on the dynamic virtual energy storage index of air conditioners and the virtual charging and discharging power of the electric vehicle cluster. The module uses a dynamic multi-objective particle swarm optimization algorithm to dynamically solve the weights of the objective function, resulting in a flexible load joint scheduling strategy for local consumption of distributed photovoltaic power.

[0175] To verify the effectiveness and rationality of the proposed method for joint scheduling of virtual energy storage for flexible loads in public buildings for on-site consumption of distributed photovoltaic power, a public building in Nanjing with a total cooling area of ​​13238 m² is used as the research object. 2 The building is 41.9m high with 11 floors above ground; the ventilation rate of the system is 430m³ / h. 3 / h, air density is 1.29kg / m³ 3 The specific heat capacity is taken as 0.995 kJ / (kg·K); the wall thickness is 240 mm, and the density is 2600 kg / m³. 3 The heat transfer coefficient of the exterior wall to the outside is taken as 4.74 W / (m²). 2 • K); The heat transfer coefficient of the wall and interior is taken as 2.8 W / (m²). 2 •K); The equivalent heat capacity coefficient of the building is taken as 150J / (m²·K); It is assumed that the maximum charging power of the charging pile in the public building is 60kW and the maximum discharging power is 20kW; The number of electric vehicles that arrive at the public building in one day is set to 50, and the rated battery capacity of each electric vehicle is 40kWh; The expected state of charge of electric vehicle users when leaving the public building is set to 0.8.

[0176] like Figure 3As shown in the figure, the virtual charging and discharging power of the air conditioners is as follows: From 5:00 to 13:00, the air conditioners are in a virtual charging state, storing excess photovoltaic power in the virtual energy storage, thus providing virtual charging power to the air conditioner cluster. As photovoltaic output increases, the virtual charging power of the air conditioners also increases. From 13:00 to 15:00, the air conditioners switch to a virtual discharging state, but due to the drop in outdoor temperature, the discharge power remains at a low level. At 17:30, the virtual discharge power of the air conditioners increases significantly, mainly due to the air conditioner set temperature rising from 25℃ to 27.5℃, the increase in the number of electric vehicles leaving, and the sharp decrease in photovoltaic output, causing the air conditioner cluster to reduce its cooling capacity and operate in virtual discharge mode. From 17:45 to 20:00, the air conditioners remain in a virtual discharge state. During this time, the outdoor temperature gradually decreases, and the operating power of the air conditioners also decreases accordingly, while the photovoltaic output gradually decreases until it reaches zero.

[0177] like Figure 4 As shown in the figure, this graph presents the virtual energy storage capacity and SOC of the air conditioner. The air conditioner's start-up time coincides with the photovoltaic (PV) start-up time, allowing the air conditioner to store PV energy using virtual energy storage during PV output. Between 7:00 and 15:30, the virtual energy storage capacity and SOC of the air conditioner are at a high level, indicating high utilization of the virtual energy storage and storage of a large amount of PV power. At 15:30, the virtual capacity of the air conditioner suddenly decreases and shows a downward trend, consistent with the analysis of the air conditioner's virtual charge and discharge power. The virtual state of charge (SOC) of the air conditioner rises rapidly during the 5:00-6:00 start-up period, reaching 0.93 around 7:00 and stabilizing. During this period, the air conditioner's SOC is high, and the adjustable margin of the virtual energy storage capacity is small. At 17:30, the virtual SOC of the air conditioner changes due to the above reasons, eventually stabilizing at around 0.6, indicating a large adjustment margin.

[0178] like Figure 5 As shown in the figure, this graph presents the virtual charging and discharging power of electric vehicles. From 0:00 to 4:45, the number of arriving electric vehicles is relatively small, and there is no photovoltaic output. Since electric vehicles generally do not have urgent travel needs during this period, virtual discharge power is used for optimization. The first peak in electric vehicle charging occurs from 6:00 to 15:00. During this time, photovoltaic output is sufficient, and electric vehicles are in virtual charging power, storing excess photovoltaic energy to supply the building when photovoltaic power is insufficient. From 15:00 to 16:30, photovoltaic output drops rapidly, and the supply of clean energy decreases sharply. Therefore, from 16:00 to 19:45, electric vehicles that have not yet left and leave later are in virtual discharge power, releasing previously stored electricity. After 20:00, most electric vehicles are in charging mode to meet the minimum SOC requirements of electric vehicles when users leave the building.

[0179] like Figure 6As shown in the figure, the virtual capacity and SOC of electric vehicles are presented. With the increase in the number of parked vehicles, electric vehicles are mainly in a virtual charging state from 5:00 to 15:00. From 16:00 to 19:45, the photovoltaic output decreases significantly, the air conditioning cluster reduces its power during this period, and the electric vehicles virtually discharge to supplement the building's power demand, resulting in a downward trend in the overall optimizable discharge capacity. After 19:00, the virtual capacity is generally low, and most electric vehicles have left the building. The overall control strategy can be seen from the changes in the virtual state of charge of the electric vehicle cluster. At 0:00, the newly connected electric vehicles have a high state of charge, which is then controlled by virtual discharge power, gradually reducing the virtual state of charge to around 0.2. After the photovoltaic system starts outputting power at 5:00, the electric vehicles switch to a virtual charging state, and the virtual state of charge gradually rises to around 0.8. At 19:30, most electric vehicles leave the building, and the remaining electric vehicles operate in a virtual discharge state to supplement the building's power demand, causing the virtual state of charge to decrease accordingly. Between 21:00 and 24:00, in order to meet users' expected power levels, the virtual state of charge of electric vehicles gradually increases.

[0180] like Figure 7 The diagram illustrates the relationship between indoor and outdoor temperatures. When the air conditioner is off, the indoor and outdoor temperatures are the same. After the air conditioner is turned on at 5:00 AM, due to the cold storage characteristics of the building envelope, the indoor temperature gradually approaches the set temperature. During the stable operation phase of the air conditioner, the indoor temperature remains essentially consistent with the set temperature. After the air conditioner is turned off at 8:00 PM, due to the cold storage characteristics of the building envelope, the indoor temperature slowly drops to the outdoor temperature.

[0181] like Figure 8 As shown, during the period from 5:00 to 6:00, the photovoltaic (PV) system is just beginning to generate power, and the base load electricity consumption is low, resulting in a PV utilization rate of approximately 0.7. From 6:00 to 16:00, PV output is sufficient, and the virtual energy storage systems for air conditioning and electric vehicles jointly absorb the PV energy, leading to a higher PV utilization rate. From 16:00 to 18:45, the PV utilization rate is essentially at full absorption, as PV output rapidly decreases during this period, allowing the building base load and the two virtual energy storage systems to fully absorb the PV energy. The PV utilization rate during other periods is around 0.9, indicating that the combined optimization of the two virtual energy storage systems can flexibly store PV resources in virtual energy storage and release them in a timely manner, fully absorbing PV resources.

Claims

1. A method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, characterized in that, Includes the following steps: (1) Considering temperature changes, a dynamic virtual energy storage model for a single air conditioner is constructed based on the dynamic virtual energy storage index of the air conditioner; the dynamic virtual energy storage index of the air conditioner includes the virtual charging and discharging power of the air conditioner, the virtual capacity of the air conditioner and the virtual state of charge of the air conditioner. (2) Based on the Monte Carlo distribution of the start-up time, continuous operation time and optimal set temperature of the air conditioners in each room of the public building, generate a Monte Carlo-based air conditioner cluster scenario and construct a dynamic virtual energy storage model for the air conditioner cluster. (3) Based on the probability density function of the arrival time of electric vehicles in public buildings, the departure time of electric vehicles from public buildings, and the initial state of charge of electric vehicles upon arrival in public buildings, generate a Monte Carlo-based electric vehicle cluster scenario and construct a dynamic virtual energy storage model for the electric vehicle cluster. (4) Based on the dynamic virtual energy storage index of air conditioners and the virtual charging and virtual discharging power of electric vehicle clusters, construct the objective function for the coordinated operation optimization of public buildings, and use the dynamic multi-objective particle swarm algorithm to dynamically solve the weight of the objective function to obtain a flexible load joint scheduling strategy for local consumption of distributed photovoltaics. In step (4), an optimization objective function for the coordinated operation of public buildings is constructed based on the dynamic virtual energy storage index of the air conditioner and the dynamic virtual energy storage index of the electric vehicle cluster. as follows: ； In the formula, 'a' represents the air conditioning operation and maintenance price, expressed in yuan / kWh; / These represent the charging / discharging prices of electric vehicles during the Lth optimization period, in yuan / kWh. The duration of an optimization period is expressed in hours (h); p(L) is the grid electricity price, expressed in yuan / kWh; c(L) is the carbon emission penalty cost for the Lth optimization period, expressed in yuan / kg. The dynamic carbon emission factor for the Lth optimization period is expressed in kg / kWh. The cost of deviating from comfort level is expressed in yuan. The photovoltaic power output is predicted for the Lth optimization period, in kW; The actual photovoltaic usage during the Lth optimization period is expressed in kW. These are the weighting coefficients for the two objectives; The power purchased by the building during the Lth optimization period is expressed in kW. Let PMV be the indoor temperature corresponding to the Lth optimization period. The PMV value corresponds to the optimal indoor temperature; T represents the number of optimization periods. The total virtual charging and discharging power of the air conditioning cluster, in kW; and These represent the virtual charging and virtual discharging power of the electric vehicle cluster during the Lth optimization period, respectively, in kW.

2. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 1, is characterized in that, In step (1), the virtual charging and discharging power of the air conditioner meets the following conditions: ； In the formula, P ac This refers to the actual operating power of the air conditioner, measured in kW; P ac,base It maintains the indoor temperature at the set temperature T. set The required basic power, in kW; P ac,ch / P ac,dis These represent the virtual charging and discharging power of the air conditioner, in kW; The virtual power capacity of the air conditioner is ； ； In the formula, t represents the virtual power capacity of the air conditioner during the Lth optimization period, in kWh; L This is the time the air conditioner is turned off, measured in hours (h). When the air conditioner is turned off during the Lth optimization period, the indoor temperature drops from the lowest indoor temperature T. min Rise to the highest indoor temperature T max The time is expressed in hours (h); C is the building's equivalent heat capacity coefficient, expressed in kJ / ℃; λ o λ represents the convective heat transfer coefficient between the outdoor environment and the exterior wall, expressed in W / (m²·℃); i The convective heat transfer coefficient between the indoor environment and the exterior walls; A0 is the total area of ​​the external heat storage material, in m2; h i Ti represents the coefficient of heat exchange between the i-th heat source and the indoor environment, in W / (m2·℃); T0 represents the outdoor temperature, in ℃; Ti represents the outdoor temperature. max The highest indoor temperature, expressed in °C; T min This is the lowest indoor temperature, expressed in °C. Air Conditioner Virtual State of Charge (SOC) ac (L) is ； ； ； In the formula, E(L) represents the virtual energy storage capacity of the air conditioner during the Lth optimization period, in kWh. The value range of T is 0~1. i The current indoor temperature is in °C. For the Lth optimization period, when the air conditioner is turned off, the indoor temperature starts from the minimum indoor temperature T. min Rise to the current indoor temperature T i The required time is in hours (h).

3. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 2, is characterized in that... In step (2), the Monte Carlo distribution of the start-up times of the air conditioners in each room of the public building is as follows: Time period A (09:00-18:00): ； Time period B (18:00-09:00 the next day): ； In the formula, The start-up time of the air conditioners in each room of a public building, in hours; Let A be the probability density function of the air conditioning start-up time of each room in a public building during time period A. Let be the probability density function of the air conditioning start-up time of each room in a public building during time period B; The Monte Carlo distribution of continuous air conditioning operation time in various rooms of a public building is as follows: Time period A (09:00-18:00): ； Time period B (18:00-09:00 the next day): ； In the formula, The continuous operating time of the air conditioners in each room of a public building, in hours; Let A be the probability density function of the continuous operation time of the air conditioner in each room of a public building during time period A. Let be the probability density function of the continuous operation time of the air conditioner in each room of a public building during time period B; The Monte Carlo distribution of the optimal set temperature for each room in a public building is as follows: ； In the formula, The optimal set temperature for air conditioning in each room of a public building, expressed in °C. ; The probability density function for setting the optimal temperature for each room in a public building.

4. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 3, is characterized in that... In step (2), the total virtual charging and discharging power of the air conditioning cluster in the dynamic virtual energy storage model of the air conditioning cluster is: ； In the formula, The total virtual charging and discharging power of the air conditioning cluster, in kW; During the Lth optimization period, the i-th air conditioner maintains the indoor temperature at the set temperature T. set The required basic power, in kW; The charging power of the i-th air conditioner during the L-th optimization period is expressed in kW. For the first L During the first optimization period i The discharge power of each air conditioner is expressed in kW; N represents the total number of air conditioners. The total virtual energy storage capacity of the air conditioning cluster is ； In the formula, This represents the total virtual energy storage capacity of the air conditioning cluster, expressed in kWh. The virtual energy storage capacity of the i-th air conditioner during the L-th optimization period is expressed in kWh. The total virtual energy storage capacity of the air conditioning cluster is ； In the formula, This represents the total virtual energy storage capacity of the air conditioning cluster, expressed in kWh. The virtual energy storage capacity of the i-th air conditioner during the L-th optimization period is expressed in kWh. The total virtual state of charge is ; In the formula, The value range for the Lth optimization period is 0-1.

5. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 4, is characterized in that... In step (3), the time it takes for the electric vehicle to arrive at the public building meets the following conditions: ； In the formula, Let be the probability density function of the time it takes for an electric vehicle to arrive at a public building; This represents the average time it takes for electric vehicles to arrive at public buildings. Let Variance be the time it takes for electric vehicles to reach public buildings. The time it takes for an electric vehicle to reach a public building, in hours. The electric vehicle must meet the following conditions for leaving the public building: ； In the formula, Let μ be the probability density function of the time it takes for an electric vehicle to leave a public building; r Let σ be the mean time it takes for electric vehicles to leave the public building. r Let Variance be the time it takes for electric vehicles to leave public buildings. The time it takes for an electric vehicle to leave a public building, in hours. The initial state of charge of the electric vehicle upon arrival at the public building meets the following conditions: ； In the formula: Let μ be the probability density function of the electric vehicle's initial state of charge upon arrival at the public building; B Let σ be the mean of the initial state of charge of the electric vehicle upon arrival at the public building. B Let Variance be the variance of the initial state of charge of the electric vehicle upon arrival at the public building. The initial state of charge of the electric vehicle upon arrival at the public building, with a value ranging from 0 to 1.

6. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 5, is characterized in that... In step (3), the virtual charging and discharging power of the electric vehicle cluster in the dynamic virtual energy storage model of the electric vehicle cluster satisfies the following conditions: ； ； ； In the formula, and These represent the virtual charging and virtual discharging power of the electric vehicle cluster during the Lth optimization period, in kW. The total power of the electric vehicle in the Lth optimization period is expressed in kW. The charging power of the j-th electric vehicle in the Lth optimization period, in kW; The discharge power of the j-th electric vehicle in the Lth optimization period, in kW; Let N represent the charging and discharging state of the j-th electric vehicle during the L-th optimization period, where 1 represents the charging state and 0 represents the discharging state. EV This represents the total number of electric vehicles; The virtual energy storage charging and discharging capacity of the electric vehicle cluster meets the following conditions: ； ； In the formula, and These represent the virtual energy storage charging and discharging capacities for the Lth optimization period, in kWh; N EV The number of electric vehicles participating in the optimization; Let J represent the state of charge of the j-th electric vehicle during the L-th optimization period. The maximum expected state of charge for electric vehicle users; E represents the minimum expected state of charge for electric vehicle users; E is the rated battery capacity of the electric vehicle. The total virtual electric capacity of electric vehicles participating in the operation optimization during the Lth optimization period. for ； The total virtual state of charge of the electric vehicle cluster is ； In the formula, Let SOC be the total virtual energy storage of electric vehicles in the Lth optimization period, with a value ranging from 0 to 1.

7. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 6, is characterized in that... In step (4), the dynamic multi-objective particle swarm optimization algorithm is used to dynamically solve the weights of the objective function. In N solutions, the formula for calculating the weight coefficients is: ； The formula for updating particle velocity is: ； In the formula, Let be the velocity of the particle in the Nth iteration. The velocity of the particle in the (N+1)th iteration calculation; and These are independent random numbers within the interval [0, 1]. This is the current optimal position for the particle; The globally optimal position; inertial constant The value decreases linearly in the range [0, 1]. Let i be the position of the i-th particle in the Nth iteration; It is a cognitive learning factor used to adjust the weight of individual experiences; This is a social learning factor used to adjust the weights of social experience.

8. The method for joint scheduling of virtual energy storage for flexible loads in public buildings for local consumption of distributed photovoltaic power, as described in claim 7, is characterized in that... The formula for updating particle positions is: ； In the formula, Let be the position of the i-th particle in the Nth iteration.

9. A joint dispatch system for virtual energy storage of flexible loads in public buildings for local consumption of distributed photovoltaic power, characterized in that, The system includes: a module for constructing a dynamic virtual energy storage model for a single air conditioner, which considers temperature changes and constructs a dynamic virtual energy storage model for a single air conditioner based on its dynamic virtual energy storage indicators; the dynamic virtual energy storage indicators for the air conditioner include virtual charging and discharging power, virtual capacity, and virtual state of charge; a module for constructing a dynamic virtual energy storage model for an aggregated air conditioner, which generates a Monte Carlo-based air conditioner cluster scenario based on the Monte Carlo distribution of the start-up time, continuous operation time, and optimal set temperature of the air conditioners in each room of the public building, and constructs a dynamic virtual energy storage model for the air conditioner cluster; and a module for constructing a dynamic virtual energy storage model for an aggregated electric vehicle cluster, which generates a Monte Carlo-based electric vehicle cluster scenario based on the probability density function of the arrival time of electric vehicles in the public building, the departure time of electric vehicles from the public building, and the initial state of charge of electric vehicles upon arrival in the public building, and constructs a dynamic virtual energy storage model for the electric vehicle cluster. The strategy solving module is used to construct an optimization objective function for the coordinated operation of public buildings based on the dynamic virtual energy storage index of air conditioners and the virtual charging and discharging power of electric vehicle clusters. The dynamic multi-objective particle swarm algorithm is used to dynamically solve the weights of the objective function to obtain a flexible load joint scheduling strategy for local consumption of distributed photovoltaic power. Based on the dynamic virtual energy storage indicators of air conditioners and electric vehicle clusters, an optimization objective function for the coordinated operation of public buildings is constructed. as follows: ； In the formula, 'a' represents the air conditioning operation and maintenance price, expressed in yuan / kWh; / These represent the charging / discharging prices of electric vehicles during the Lth optimization period, in yuan / kWh. The duration of an optimization period is expressed in hours (h); p(L) is the grid electricity price, expressed in yuan / kWh; c(L) is the carbon emission penalty cost for the Lth optimization period, expressed in yuan / kg. The dynamic carbon emission factor for the Lth optimization period is expressed in kg / kWh. The cost of deviating from comfort level is expressed in yuan. The photovoltaic power output is predicted for the Lth optimization period, in kW; The actual photovoltaic usage during the Lth optimization period is expressed in kW. These are the weighting coefficients for the two objectives; The power purchased by the building during the Lth optimization period is expressed in kW. Let PMV be the indoor temperature corresponding to the Lth optimization period. is the PMV value corresponding to the optimal indoor temperature; T is the number of optimization periods; The total virtual charging and discharging power of the air conditioning cluster, in kW; and These represent the virtual charging and virtual discharging power of the electric vehicle cluster during the Lth optimization period, respectively, in kW.

Images