Household micro-grid energy optimization scheduling method considering demand response based on time-of-use electricity price

Abstract

The invention discloses a household micro-grid energy optimization scheduling method considering demand response based on time-of-use electricity price. The method comprises the following steps: constructing a household micro-grid model considering various demand loads; considering a discharge depth constraint and an energy storage cycle number constraint, and constructing a full-cycle battery life model; for uncertainty of new energy output of the household micro-grid, a Latin hypercube sampling method is combined with a Kantorovich scene reduction method to generate a photovoltaic output typical scene; and constructing a micro-grid energy optimization scheduling model and solving by taking the minimum family operation cost, energy purchase cost and daily loss cost of the service life of an energy storage battery as a target and combining constraint conditions. According to the method, the power consumption cost of the household micro-grid can be effectively reduced, and the consumption level of new energy is improved. The technical support is provided for the safe operation between the distributed photovoltaic system and the power grid while the home user actively participates in the demand response scheduling.

Description

Technical Field

[0001] The present invention relates to the field of microgrid optimal scheduling, and in particular to a method for optimizing the energy scheduling of a household microgrid considering demand response under time-of-use electricity prices. Background Art

[0002] With the continuous improvement of the living standards of Chinese residents, the types and quantities of household electricity loads have increased significantly. Household electricity loads show a trend of diversification and growth. The growing electricity demand of residents not only poses higher requirements for the operation of traditional power grids but also presents new challenges for the optimization and management of household energy usage patterns. Against this background, with the development of distributed photovoltaic systems, they provide diversified options for the energy supply methods of household users. However, how to achieve efficient coordination among household loads, the power grid, as well as photovoltaics and energy storage to ensure the stable operation of the household microgrid has become one of the hot issues in current research.

[0003] Literature [1]: Zhou Lei, Li Yang. Modeling and Optimal Operation of Household Loads Based on a Home Energy Management System in a Time-of-Use Electricity Price Environment [J]. Power System Technology, 2015. By considering multiple types of household flexible loads and combining photovoltaic and energy storage devices, an optimal scheduling strategy for different household devices in a time-of-use electricity price environment was established. The simulation results show that reasonable scheduling of household flexible loads can effectively reduce the electricity purchase cost of users, but it does not consider the impact of the uncertainty of distributed energy output on the microgrid.

[0004] Considering that the randomness and uncertainty of photovoltaic output will have an adverse impact on the energy scheduling of the microgrid, thereby reducing the operation efficiency and scheduling reliability of the system, many studies have been devoted to proposing effective algorithms to quantify and alleviate this problem. Literature [2]: Duan Junhong, Liang Chen, Li Yaxin, etc. Optimal Scheduling Strategy for Residential Energy Storage and Load Considering Photovoltaic Uncertainty [J]. Power Demand Side Management, 2024. An affine algorithm was used to quantify the uncertainty of photovoltaic output, and an optimal scheduling model for energy storage and load with the goals of minimizing the daily electricity cost of residents and maximizing comfort was established. Through simulation verification, the proposed algorithm can fully consider the uncertainty of photovoltaic output and effectively improve the accuracy of the scheduling model. Literature [3]: Zhang Yusen, Kong Xiangyu, Sun Bowei, etc. Multi-Time Scale Home Energy Management Optimization Strategy Based on Power Demand Response [J]. Power System Technology, 2018. The output of the energy storage was continuously adjusted according to the real-time output of the photovoltaic to reduce the error between the predicted value and the actual value of the load and photovoltaic output, thereby improving the scheduling accuracy. Although the above studies can effectively reduce the uncertainty of photovoltaic output, they cannot fully depict the randomness of new energy output.

[0005] As an important part of the microgrid, the battery has shown significant advantages in improving the operating economy of the microgrid because it can effectively suppress the fluctuations of new energy. Literature [4]: Wang Laizeng, Feng Lu, Zhang Xinjing. Research on the Technical Economy of Wind-Solar-Storage System Considering Time-of-Use Electricity Price [J]. Thermal Power Generation, 2024. On the premise of meeting the reliable power supply to the load, an optimization model of the source-network-load-storage system is established with the goal of the optimal system economy for solution. The analysis results show that the battery stores energy at low electricity prices and releases energy at high electricity prices, thus avoiding users from purchasing electricity from the grid at high electricity prices, and further reducing the electricity purchase cost of users from the grid. Literature [5]: Liu Jiaojiao, Wang Zhijie, Yuan Jianhua, etc. Research on Optimal Scheduling of Wind-Solar-Storage Microgrid Based on PSO Algorithm [J]. East China Electric Power, 2014. Taking the minimum operating cost of the microgrid as the objective function and considering the full charge and discharge constraints of the battery at the same time, the particle swarm algorithm is used to solve the scheduling model. Compared with the traditional scheduling method, this method significantly improves the operating economy of the microgrid. Previous studies have shown that the introduction of batteries in the microgrid can effectively improve the operating economy of the microgrid, but the modeling of the battery is too idealized, and the operation and maintenance costs and depreciation costs of the battery use are not considered.

[0006] To sum up, on the basis of considering the uncertainty of photovoltaic output and the battery depreciation problem, an optimal scheduling model of a microgrid based on distributed photovoltaic equipment, energy storage equipment, and household demand response is proposed. To more accurately depict the stochastic characteristics of new energy output, the Latin hypercube sampling method combined with the Kantorovich scenario reduction method is introduced to generate typical scenarios of photovoltaic output, thereby improving the adaptability and robustness of the scheduling model. At the same time, in view of the loss problem of the battery during long-term operation, for this reason, a full-life-cycle battery model is constructed to fully consider the service life and depreciation problem of the battery during the optimal scheduling process. Under the time-of-use electricity price mechanism, the effectiveness and feasibility of the proposed model are verified through simulation examples. The results show that the model can effectively improve the economy and operating stability of the microgrid. Summary of the Invention

[0007] The present invention proposes an optimal scheduling method for the energy of a household microgrid considering demand response under time-of-use electricity price, which can effectively reduce the electricity cost of the household microgrid and improve the consumption level of new energy. While users at home actively participate in demand response scheduling, it also provides technical support for the safe operation between the distributed photovoltaic system and the power grid.

[0008] The technical solution adopted by the present invention is as follows:

[0009] The optimal scheduling method for the energy of a household microgrid considering demand response under time-of-use electricity price includes the following steps:

[0010] Step 1: Construct a household microgrid model considering various demand loads;

[0011] Step 2: Considering the depth of discharge constraint and the energy storage cycle number constraint, construct a full-cycle battery life model;

[0012] Step 3: Aiming at the uncertainty of new energy output in the household microgrid, use the Latin hypercube sampling method combined with the Kantorovich scenario reduction method to generate typical scenarios of photovoltaic output;

[0013] Step 4: Taking the minimum of the household operation cost, the energy purchase cost, and the daily loss cost of the energy storage battery life as the goal, combined with the constraint conditions, construct a microgrid energy optimal scheduling model and solve it.

[0014] In the above-mentioned Step 1, the household microgrid model includes:

[0015] 1) Distributed photovoltaic model:

[0016] Photovoltaic power generation is mainly related to the light intensity, its own physical parameters, and the battery temperature. Its power generation output characteristics are significantly non-linear, and its output power model is as follows:

[0017]

[0018] In formula (1): P pv (t) is the photovoltaic power generation power at time t; P pv,s is the maximum output power under the standard test when the light intensity reaches 1000 kw / m 2 , and the ambient temperature is 25°C; k is the temperature coefficient, taking -0.005°C; T j,stc is the reference temperature; T j t and NOCT are the current working temperature and the normal operating temperature of the photovoltaic respectively; G t is the light intensity at time t.

[0019] 2) Energy storage battery model:

[0020] The energy storage battery can smooth the output of distributed power sources, improve the power supply quality, store electric energy during the low grid valley period, and release electric energy during the peak period to achieve peak shaving and valley filling. Its mathematical model is as follows:

[0021]

[0022] In formula (2): SOC(t) is the energy storage battery capacity at time t; η in and η out are the charge and discharge efficiencies respectively; and are the charge and discharge efficiencies respectively; and are the upper limit values of the charge and discharge powers respectively; δin and δ out are 0-1 variables of the charging and discharging efficiency respectively, which can constrain the charging and discharging energy at the same moment; SOC max and SOC min are the upper and lower limits of the state of charge respectively; SOC(t - 1) is the energy storage battery capacity at time t - 1; △t represents the scheduling duration, taking 1h; SOC(0) represents the initial capacity of the energy storage battery; SOC(24) represents the final capacity of the energy storage battery;

[0023] 3) Demand response load model, which includes rigid load, shiftable load, transferable load, and reducible load;

[0024] ①. The rigid load is the load that guarantees the basic electricity demand of residents. Its running time and power are fixed and non - schedulable. Its model is as follows:

[0025]

[0026] In formula (3): P RL (t) is the power consumed by the rigid load in the time period; k is the number of rigid loads; is the rated power of load i; is the operating state of load i, represented by a 0 - 1 variable for off and on.

[0027] ②. The shiftable load refers to the load whose running time can be shifted within the acceptable time period of the user and cannot be interrupted once it starts running. Its model is as follows:

[0028]

[0029] In the above formula: is the power consumed by the shiftable load i in the t time period; is the rated power of load i; is the operating state of the shiftable load in the t time period; are the starting and ending time periods when the load can run respectively; is the number of consecutive time periods of the load; t represents the current time period.

[0030] ③. The interruptible load refers to the load that can run intermittently with the minimum running time within the time period allowed by the user and can be switched on and off at will while ensuring the total running time remains unchanged. Its model is as follows:

[0031]

[0032] In the above formula: is the total power consumed by the interruptible load in the time period; is the rated power of load i; The operating state of load \(i\) at time period \(t\), represented by 0 - 1 for off and on; They are respectively the start and end time periods when the load can operate; Is the number of consecutive time periods of the load.

[0033] ④. The load that can be curtailed mainly refers to the temperature - controlled load. Users control the operating power of the temperature - controlled load, so that the temperature approaches the user's desired temperature. Taking the refrigeration of an air - conditioning system as an example, its model can be expressed as:

[0034]

[0035] In Equation (10): \(R\) represents the room thermal resistance; \(C\) is the room heat capacity; \(T\) in (t + 1) represents the indoor temperature at time period \(t+1\); \(T\) in (t) represents the indoor temperature at time period \(t\); \(P\) cut (t) represents the power of the load that can be curtailed at time period \(t\); \(T\) out (t) represents the outdoor temperature at time period \(t\); Represents the exponential decay factor, reflecting the influence degree of the power of the load that can be curtailed and the outdoor temperature on the indoor temperature at time period \(t + 1\).

[0036] In step 2,

[0037] 2.1: Daily cycle number constraint. By restricting the daily cycle number during the operation of the energy storage battery, the service life of the energy storage battery is improved. The daily cycle number constraint model can be expressed as:

[0038]

[0039] In Equation (11): \(S\) bat (t) is the daily cycle state variable of the energy storage battery. When the energy storage battery switches from charging to discharging, it is recorded as 1, otherwise 0; \(U\) bat (t) is the charge - discharge state of the energy storage battery at time \(t\), 1 represents charging, 0 represents discharging; \(N\) day Represents the daily cycle number, and its value is in the interval \(\{0,1,2,\cdots,12\}\); \(U\) bat (t - 1) represents the charge - discharge state of the energy storage battery at time \(t-1\); 2.2: Depth of discharge constraint. By linearly segmenting the depth of discharge to restrict the discharge interval of the energy storage battery, its model can be expressed as:

[0040]

[0041] In Equation (12): \(dod\) d (t) is the depth of discharge of the \(d\) - th segment of the energy storage battery at the \(t\) - th time period; \(dod\) d 、 Are respectively the upper and lower limits of the depth of discharge of the \(d\) - th segment; \(g\)d (t) represents the binary state variable of the section where the depth of discharge of the energy storage battery is located; d represents the section where the depth of discharge of the energy storage battery is located.

[0042] In step 2, the full-cycle battery life model refers to effectively extending its service life by restricting the daily cycle times and depth of discharge of the energy storage battery. The full-cycle battery life model includes equations (11) to (12).

[0043] In step 3, using the typical day of wind and light as the original data, the Latin hypercube sampling method is adopted to generate multiple wind and light output data that conform to the typical day of wind and light; secondly, the scenario reduction method using the Kantorovich distance is used to select the wind and light output data that most conforms to the actual situation as the wind and light data input for microgrid dispatching, which includes:

[0044] S31. Generation of Latin hypercube sampling scenarios:

[0045] Assume that N random variables K1, K2, …, K N , K1, K2, …, K N respectively represent different generated random variables. Among them, K N 's cumulative probability distribution function F N is:

[0046] F N = F N (K N ) (13);

[0047] In equation (13): F N (K N ) represents the cumulative probability distribution function of the random variable K N .

[0048] Let the sampling scale be R, then the sampling steps of K N are as follows:

[0049] S311. Divide the distribution curve of F N into R intervals, each interval has the same range, which is all

[0050] S312. Extract a number in each interval. For example, the sampling point K Ni in the i-th interval, its cumulative probability P Ni is:

[0051]

[0052] In equation (14): r i is a random number in the interval [0, 1].

[0053] S313. Take PNi Substitute into the function to obtain the sampled value K in the corresponding interval Ni :

[0054]

[0055] In Equation (15): represents the corresponding cumulative probability P Ni of the sampled value

[0056] S314. Sample R times through the above method, and R sampled values of K N will be generated

[0057] S315. Generate an N×R-dimensional matrix and randomly sort each row to generate R scenarios

[0058] S32. Scenario reduction method combined with Kantorovich distance

[0059] Assume that the number of photovoltaic output power prediction scenarios generated by Latin hypercube sampling is N, and the number of reduced scenarios is n. Then the steps are as follows

[0060] S321. Initialization: The probability value of each photovoltaic output power prediction scenario is p i represents the probability value of the photovoltaic output power prediction scenario, and the initial number of reduced scenarios is n * =N

[0061] S322. Calculate the Kantorovich distance D i , s j ) between each scenario (s k (s i , s j ), s i represents the scenario i to be compared; s j represents the scenario j to be compared

[0062] Then the Kantorovich distance of each scenario is

[0063]

[0064] In Equation (16): D k (s i , s j ) represents the Kantorovich distance between scenario i and scenario j; L i,t represents the wind and light output value of scenario i at time t; L j,t represents the wind and light output value of scenario j at time t; S represents the total number of scenarios

[0065] S323. Select the scenario sk The scenario s with the minimum distance r , and calculate the product of the Kantorovich distance and the scenario probability, which can be denoted as:

[0066]

[0067] In Equation (17): represents the probability of the occurrence of scenario r.

[0068] S324. Repeat step S323 for each scenario, and then select the scenario with the minimum value and denote it as scenario d, and delete this scenario. At the same time, update the reduced number of scenarios n * = n * - 1, then the probability value of scenario r can be updated to p r = p r + p d .

[0069] S325. Repeat steps S322 - S324 to obtain the final number of scenarios n * = n.

[0070] In step 4, taking demand response as the core of user energy optimization, aiming at minimizing the home operation cost, energy purchase cost, and daily loss cost of the energy storage battery life, combined with multiple constraint conditions, a microgrid energy optimization scheduling model is constructed for solution, which includes:

[0071] (1). Objective function operation cost:

[0072] Comprehensively considering the energy purchase and sale costs of the home microgrid, the operation and maintenance costs of equipment, and the daily loss costs of the energy storage battery, and taking the minimization of the total operation cost F as the scheduling objective, as shown in Equation (18) below:

[0073]

[0074] In Equation (18): F nent 、F op 、F bat are respectively the energy purchase and sale costs of the home microgrid, the operation and maintenance costs of equipment, and the daily loss costs of the energy storage battery; are respectively the electricity purchase price and the electricity sale price at time t; Z in F op represents the type of equipment, specifically the photovoltaic and energy storage equipment here; λ i is the unit output maintenance parameter of the i-th equipment; P t i is the output of the i-th equipment at time t; E bat,max is the configured battery energy storage capacity; g d(t) is a piecewise 0-1 variable of the depth of discharge, and taking 1 means that the depth of discharge of the energy storage battery at time t is in the d-th segmented interval. P buy represents the purchased electric power; P sell represents the sold electric power; T represents the total dispatching duration, specifically referring to the time of a day; B s (N day ) represents the binary state variable of the daily cycle times.

[0075] (2). The constraint conditions of the objective function include:

[0076] Electric power constraint: The electric power balance constraint formula of this household microgrid is as follows:

[0077]

[0078] In formula (19): P pv (t), P buy (t), P sell (t) are the photovoltaic output, purchased electricity quantity, and sold electricity quantity at time t respectively;

[0079] P in (t), P out (t) are the charging power and discharging power of the energy storage battery at time t respectively;

[0080] are the power of the i-th rigid load, the i-th shiftable load, the i-th interruptible load, and the i-th curtailable load at time t respectively;

[0081] N1, N2, N3, N4 refer to the number of rigid loads, the number of shiftable loads, the number of interruptible loads, and the number of curtailable loads respectively;

[0082] refer to the purchase and sale electricity limits respectively;

[0083] δ buy 、δ sell are 0-1 variables during power purchase and sale respectively, which restrict that the microgrid cannot sell and purchase electricity simultaneously at the same time. The microgrid energy optimization scheduling model refers to a mathematical optimization framework, aiming to minimize the household electricity cost through coordinating the scheduling of distributed energy sources (such as photovoltaic), energy storage devices (such as storage batteries), and household loads under the time-of-use electricity price and demand response mechanism. Its core is to balance the goals such as economy and equipment life by reasonably allocating the time sequence of energy production, storage, and consumption, and it includes formulas (18) to (19).

[0084] The solution of the microgrid energy optimization scheduling model includes the following steps:

[0085] Step 1: Data Preparation: Collect household load data, historical PV output data, time-of-use electricity price information, and battery parameters. Step 2: Scenario Generation: Use Latin Hypercube Sampling to generate a large number of PV output scenarios that cover the possible fluctuation range; apply the Kantorovich scenario reduction method to screen out representative typical scenarios based on probability distance, reducing the computational complexity.

[0086] Step 3: Model Solving: Use the Yalmip toolbox to establish an optimization model in MATLAB, call the Cplex solver for mixed-integer linear programming to solve, and the output result is the optimal scheduling strategy for each period, including the battery charge and discharge plan, the start and stop times of flexible loads, and the grid power purchase / sale decision;

[0087] Step 4: Simulation Verification and Effect Evaluation: Input actual household load and PV output data and run the optimization model.

[0088] The present invention relates to a method for optimizing the energy scheduling of a household microgrid considering demand response under time-of-use electricity prices, and the technical effects are as follows: 1) The present invention introduces an energy storage battery considering life loss into the household microgrid, making the microgrid more in line with reality. After introducing the daily life loss of the energy storage battery, the service life of the energy storage battery can be effectively extended.

[0089] 2) Considering the uncertainty of PV output, the present invention uses Latin Hypercube Sampling to generate PV scenarios and combines the Kantorovich distance for scenario reduction, reducing the uncertainty of PV output and providing a data basis for the reliability analysis of the household microgrid.

[0090] 3) The present invention introduces various demand responses into the household microgrid and optimizes the scheduling according to the time-of-use electricity price information for different demand responses, effectively reducing the electricity cost of microgrid users. Description of the Drawings

[0091] The following further describes the present invention in conjunction with the drawings and embodiments:

[0092] Figure 1 It is the structure diagram of the household microgrid.

[0093] Figure 2 It is the energy scheduling flowchart of the household microgrid.

[0094] Figure 3 It is the scenario generation diagram of PV.

[0095] Figure 4 It is the scenario reduction diagram of PV.

[0096] Figure 5 It is the probability diagram of each scenario after reduction.

[0097] Figure 6 It is the electrical balance power diagram.

[0098] Figure 7 It is the user-side flexible electrical load distribution diagram before optimization.

[0099] Figure 8 It is the user-side flexible electrical load distribution diagram after optimization.

[0100] Figure 9 It is the indoor temperature curve after the dispatch of the load that can be curtailed.

[0101] Figure 10 It is the energy storage change diagram of the energy storage battery. Specific implementation manner

[0102] A method for optimizing the energy dispatch of a household microgrid considering demand response based on time-of-use electricity price is proposed. A household microgrid model is constructed, which includes distributed photovoltaic units and energy storage devices, and considers various demand loads, such as shiftable loads, curtailable loads, and shiftable loads. Considering the constraints such as the life loss, depth of discharge, and number of cycles of the energy storage device, a full-cycle battery life model is established to extend the service life of the energy storage device. Aiming at the uncertainty of new energy in the household microgrid, the Latin hypercube sampling method combined with the Kantorovich scenario reduction method is used to generate typical scenarios of photovoltaic power output. Taking the energy purchase cost, equipment operation cost, and daily loss cost of energy storage life as the objective function and considering various constraints, the model is solved using a solver.

[0103] The structure diagram of the household microgrid studied in the present invention is as Figure 1 shown. As can be seen from Figure 1 , the household microgrid mainly includes distributed photovoltaic units, energy storage batteries, the power grid, rigid loads, and various demand responses.

[0104] The energy dispatch flow chart of the household microgrid is as Figure 2 shown. Among them, the main power supply of the microgrid comes from distributed photovoltaic units and the superior power grid. In order to reduce the daily electricity cost of users, a coordinated optimization strategy for multiple demand responses is constructed based on the time-of-use electricity price mechanism. Through the dynamic regulation of demand response resources such as shiftable loads, transferable loads, and curtailable loads, the minimum electricity cost is achieved. In order to improve the utilization rate of new energy and reduce the electricity purchase volume of users, an energy storage battery is introduced. Through the charge and discharge strategy of the energy storage battery, the time-domain transfer of photovoltaic power generation is realized, the new energy consumption rate is improved, and the electricity purchase volume of users is reduced.

[0105] The case of the present invention uses the matlab2023b software for simulation verification. A typical household electrical load is used for microgrid dispatch, and the household electrical load is shown in Table 1.

[0106] Table 1 Typical household electrical load of a certain household

[0107]

[0108] For the uncertainty of photovoltaic power output, the present invention first assumes that the photovoltaic power follows a normal distribution N(μ,δ 2 ), where μ is the expected value of the predicted value and δ is the percentage of its fluctuation. Then, Latin hypercube sampling is used to generate 1000 groups of photovoltaic power output scenarios that comply with the probability distribution constraints. As Figure 3 shown, it can be found from Figure 3 that the 1000 groups of generated photovoltaic scenarios all conform to the trend of the "inverted U-shaped" curve of photovoltaic power output.

[0109] The scenario reduction technology considering the Kantorovich distance is used to process this scenario for scenario reduction, and the final typical daily scenarios of photovoltaic power output are reduced to 5, as Figure 4 shown. Finally, according to the probability corresponding to each scenario, as Figure 5 shown, the scenario with the highest occurrence probability, that is, Scenario 3, is selected as the typical daily photovoltaic power output curve.

[0110] The time-of-use electricity price parameters and energy storage device parameters involved in the present invention are shown in Table 2 and Table 3 respectively.

[0111] Table 2 Time-of-Use Electricity Price Table

[0112]

[0113] Table 3 Energy Storage Battery Parameters

[0114]

[0115] To verify the effectiveness of the household microgrid energy optimization scheduling model considering demand response based on time-of-use electricity price constructed by the present invention, the following four scenarios are set for comparative analysis.

[0116] Scenario 1: Demand response does not participate in scheduling, all photovoltaic power is sold, and there is no energy storage device.

[0117] Scenario 2: Demand response participates in scheduling, all photovoltaic power is sold, and there is no energy storage device.

[0118] Scenario 3: Demand response participates in scheduling, photovoltaic power is self-consumed, and the surplus power is fed into the grid, and there is no energy storage device.

[0119] Scenario 4: Demand response participates in scheduling, photovoltaic power is self-consumed, the surplus power is fed into the grid, and the energy storage device participates in scheduling.

[0120] The optimization results of the system under different scenarios are shown in Table 4.

[0121] Table 4 Optimization Results of the System under Different Scenarios

[0122]

[0123] From the overall analysis, compared with scenario 2, after the introduction of demand response, the total cost of scenario 1 dropped from 33.60 yuan to 21.09 yuan, a decrease of about 37.3%, indicating that demand response effectively reduced the cost of electricity purchase by dynamically adjusting the power load; compared with scenario 2, the photovoltaic power generation was changed from all sales to self-use, and the power sales revenue was reduced to 0, but the power purchase cost dropped from 23.73 yuan to 17.82 yuan, a decrease of about 24.9%. This shows that the self-use of photovoltaic power reduced the purchase of electricity from the power grid. Although the power sales revenue disappeared, the overall cost was further reduced; compared with scenario 3, after adding energy storage, the total cost of scenario 4 dropped from 17.82 yuan to 16.70 yuan, a decrease of about 6.3%. Energy storage further optimizes the cost of electricity purchase by storing electricity during low-price periods and releasing it during high-price periods. Analyzing from the demand response aspect, demand response significantly reduces the cost of electricity purchase by adjusting various demand response loads to the period of low electricity price, reducing the high-priced electricity purchase during peak hours; the strategic adjustment of photovoltaic power generation has reduced the total cost from 21.09 yuan to 17.82 yuan, a decrease of 15.5%. The benefit of photovoltaic power self-use is significantly higher than the income from electricity sales. Therefore, although the income from electricity sales has disappeared, the total cost has further decreased; with the addition of energy storage equipment, the total cost has been reduced from 17.82 yuan to 16.70 yuan, a decrease of 6.3%. The energy storage equipment is charged during the period of low electricity price and discharged during the peak period, replacing the high-priced electricity purchase and optimizing the cost of electricity purchase.

[0124] Analysis of the scheduling results of scenario 4: The scheduling results of scenario 4 are as follows: Figures 6 to 10 As shown. Figure 6 The power balance analysis shown in the figure shows that after coordinated dispatch optimization, the operation of the microgrid presents significant spatiotemporal characteristics: during the high electricity price period (corresponding to the peak period of photovoltaic output during the day), the demand for electricity load decreases. During this period, most of the electricity is supplied by the photovoltaic system and energy storage batteries, and only a small amount of electricity needs to be purchased from the distribution network; during the low electricity price period, the electricity load rises sharply, and the proportion of power supply from the distribution network increases significantly. This differentiated feature stems from the coupling effect of two dimensions: first, the demand response mechanism guides the original demand response load in the high electricity price period to the low electricity price period through price signals; second, the photovoltaic system adopts the dynamic strategy of "self-generation and self-use + surplus power access" during the day to achieve the decoupling optimization of photovoltaic output and load curve - when the photovoltaic output exceeds the local load, the surplus power is fed back to the grid at a price of 0.34 yuan / kWh, which reduces the net power purchase of the microgrid during the high electricity price period. This spatiotemporal energy management strategy effectively improves the system economy and verifies the engineering value of flexible resource coordinated dispatch.

[0125] like Figure 7 , Figure 8As shown, the optimization comparison of flexible electrical loads on the user side of the microgrid shows that through the coordinated scheduling strategy, both the spatio-temporal distribution of the system load and the operating economy have been significantly improved. Before optimization, the load peak during high electricity price periods remained high due to concentrated demand. Through the transfer and translation of flexible loads and the dynamic regulation of load curtailment, the load curve shows a smoothed characteristic, and the peak-to-valley difference has been significantly narrowed. At the same time, the load restructuring strategy promotes the transformation of the operation mode: during high electricity price periods, through the coordinated scheduling of on-site PV consumption and flexible loads, the dependence on the distribution network is reduced; while the adaptive adjustment of temperature control equipment further compresses the spatio-temporal proportion of ineffective energy consumption. This optimized scheduling method provides methodological support for building an economical and efficient microgrid energy management system.

[0126] Figure 9 The indoor temperature curve after scheduling shows that, under the constraint of a significant reduction in load curtailment, the indoor temperature curve still remains stable within the comfortable range, verifying the accurate control of thermal inertia by the adaptive adjustment algorithm of the temperature control equipment, effectively suppressing ineffective energy consumption and reducing the electricity cost. At the same time, it reduces ineffective energy consumption and the electricity purchase volume of users.

[0127] Figure 10 This is the energy storage change of the energy storage battery, which realizes a deep coupling with the electricity price signal - electrical energy is stored during low electricity price periods, and strategic discharging supports load demand during high electricity price periods. At the same time, by introducing the daily cycle number constraint and discharge depth constraint of the energy storage battery, the charge and discharge frequency of the energy storage battery is effectively reduced, its service life is extended, and the operation and maintenance cost of the energy storage battery is reduced.

Claims

1. A household microgrid energy optimization scheduling method considering demand response under time-of-use electricity prices, characterized by The following steps are involved: Step 1: Construct a household microgrid model considering multiple demand loads; Step 2: Considering the constraints of discharge depth and energy storage cycle number, a full-cycle battery life model is constructed; Step 3: In view of the uncertainty of the renewable energy output of household microgrids, the Latin hypercube sampling method combined with the Kantorovich scenario reduction method is used to generate typical scenarios of photovoltaic output; Step 4: With the goal of minimizing household operating costs, energy purchase costs, and daily loss costs of energy storage batteries, combined with constraints, a microgrid energy optimization scheduling model is constructed and solved.

2. The method for optimizing energy dispatching of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1 is characterized in that: In step 1, the household microgrid model includes: 1) Distributed photovoltaic model: The photovoltaic power generation output power model is as follows: In formula (1): P pv (t) is the photovoltaic power generation power at time t; P pv,s The light intensity reaches 1000kw / m 2 , the maximum output power under the standard test with an ambient temperature of 25°C; k is the temperature coefficient; T j,stc is the reference temperature; and NOCT are the current working temperature and normal operating temperature of the photovoltaic system respectively; G t is the light intensity at time t; 2) Energy storage battery model: The mathematical model of energy storage battery is as follows: In formula (2), SOC(t) is the capacity of the energy storage battery at time t; η in and η out are the charge and discharge efficiency respectively; and are the charge and discharge efficiency respectively; and are the upper limits of charge and discharge power respectively; δ in With δ out are 0-1 variables of charge and discharge efficiency, respectively, constraining the charge and discharge energy at the same time; SOC max With SOC min are the upper and lower limits of the state of charge respectively; SOC(t-1) is the capacity of the energy storage battery at time t-1; △t represents the scheduling duration; SOC(0) represents the initial capacity of the energy storage battery; SOC(24) represents the final capacity of the energy storage battery.

3. The method for optimizing energy dispatching of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1 is characterized in that: Also includes: 3) Demand response load model, which includes rigid load, movable load, transferable load and curtailable load; ①. The rigid load model is expressed as follows: In formula (3): P RL (t) is the power consumed by the rigid load in the time period; k is the number of rigid loads; is the rated power of load i; is the operating status of load i, with a 0-1 variable representing shutdown and operation; ②. The model of the translatable load is expressed as follows: In the above formula: is the power consumed by the movable load i in period t; is the rated power of load i; is the operating state of the movable load in period t; They are the start and end periods of the load operation respectively; is the number of load duration periods; t represents the current period; ③. The interruptible load model is expressed as follows: In the above formula: The total power consumed by the interruptible load during the time period; is the rated power of load i; is the operating status of load i in period t, with 0-1 representing shutdown and operation; They are the start and end periods of load operation respectively; is the number of load duration periods; ④. The reducible load refers to the temperature control load. The user controls the operating power of the temperature control load. For the cooling of the air conditioning system, its model is expressed as: In formula (10), R represents the room thermal resistance; C represents the room heat capacity; T in (t+1) represents the indoor temperature during the period t+1; T in (t) represents the indoor temperature during period t; P cut (t) represents the load power that can be reduced during period t; T out (t) represents the outdoor temperature during period t; It represents the exponential decay factor, which reflects the influence of the load power that can be reduced and the outdoor temperature on the indoor temperature during the period t+1.

4. The method for optimizing energy dispatch of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1, characterized in that: In step 2, the daily cycle number constraint constrains the daily cycle number during the operation of the energy storage battery to increase the service life of the energy storage battery. The daily cycle number constraint model is expressed as: In formula (11): S bat (t) is the daily cycle state variable of the energy storage battery. When the energy storage battery switches from charging to discharging, it is recorded as 1, otherwise it is 0; U bat (t) is the charge and discharge state of the energy storage battery at time t, 1 represents charging, and 0 represents discharging; N day represents the number of daily cycles, and its value is in the interval {0,1,2,…,12}; U bat (t-1) represents the charging and discharging status of the energy storage battery at time t-1.

5. The method for optimizing energy dispatching of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 4 is characterized in that: The discharge depth constraint constrains the discharge interval of the energy storage battery by piecewise linearizing the discharge depth. Its model is expressed as: In formula (12): dod d (t) is the discharge depth of the energy storage battery in the tth period and the dth segment; dod d , are the upper and lower limits of the discharge depth of the th segment respectively; g d (t) represents the binary state variable of the segment where the energy storage battery discharge depth is located; d represents the segment where the energy storage battery discharge depth is located.

6. The method for optimizing energy dispatch of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1, characterized in that: In step 3, a typical wind-solar day is used as the original data, and a Latin hypercube sampling method is used to generate multiple wind-solar output data that meet the typical wind-solar day for typical scenarios; secondly, the scene reduction method of Kantorovich distance is used to select the wind-solar output data that best meets the actual situation as the wind-solar data input for microgrid scheduling.

7. The method for optimizing energy dispatching of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1, characterized in that: Step 3 includes the following steps: S31. Latin Hypercube Sampling Scenario Generation: Suppose there are N random variables K1, K2, …, K N ,K1,K2,…,K N Respectively represent different random variables generated; among them, K N The cumulative probability distribution function F N for: F N =F N (K N ) (13); In formula (13): F N (K N ) represents the random variable K N The cumulative probability distribution function of ; Assume the sampling scale is R, then K N The adoption steps are as follows: S311. F N The distribution curve is divided into R intervals, each interval has the same range, both S312. Extract a number in each interval, the sampling point K of the i-th interval Ni , its cumulative probability P Ni for: In formula (14): r i is a random number in the interval [0,1]; S313.P Ni Bring in Function, get the corresponding interval sampling value K Ni : In formula (15): Indicates the corresponding cumulative probability P Ni The sampling value of S314. Sampling R times by the above method will generate K N R sample values ​​of ; S315. Generate an N×R dimensional matrix, and randomly sort the rows to generate R scenarios; S32. Scene reduction method combined with Kantorovich distance: Assuming that the number of PV output prediction scenarios generated by Latin hypercube sampling is N, and the number of reduced scenarios is n, the steps are as follows: S321. Initialization: The probability value of each photovoltaic output prediction scenario is p i Represents the probability value of the photovoltaic output prediction scenario, and the initial reduction scenario number is n * =N; S322. Calculate each scene (s i ,s j )'s Kantorovich distance D k (s i ,s j ), s i Indicates the scene to be compared i; s j represents the scene j to be compared; Then the Kantorovich distance of each scene is: In formula (16): D k (s i ,s j ) represents the Kantorovich distance between scene i and scene j; L i,t represents the wind and solar power output value of scene i in period t; L j,t It indicates the wind and solar power output value of scene j in time period t; S indicates the total number of scenes; S323. Selection and scenes k The scene with the smallest distance r , and calculate the product of Kantorovich distance and scene probability, which can be recorded as: In formula (17): represents the probability of scene r appearing; S324. Repeat step S323 for each scene, and then select The smallest scene is recorded as scene d, and the scene is deleted, and the number of reduced scenes n is updated at the same time * =n * -1, then the probability value of scene r can be updated to p r =p r +p d ; S325. Repeat steps S322 to S324 to obtain the final number of scenes n * =n.

8. The method for optimizing energy dispatch of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 1, characterized in that: In step 4, demand response is taken as the core of user energy optimization, and the goal is to minimize household operating costs, energy purchase costs and daily loss costs of energy storage batteries. Combined with multiple constraints, a microgrid energy optimization scheduling model is constructed for solution.

9. The method for optimizing energy dispatch of a household microgrid based on time-of-use electricity prices and considering demand response according to claim 8, characterized in that: Step 4 includes: (1) Objective function running cost: The energy purchase and sales costs of the household microgrid, the operation and maintenance costs of the equipment, and the daily loss costs of the energy storage battery are comprehensively considered, and the minimization of the total operating cost F is taken as the scheduling goal, as shown in the following formula (18): In formula (18): F nent 、F op 、F bat They are the electricity purchase and sales costs of the household microgrid, the operation and maintenance costs of the equipment, and the daily loss costs of the energy storage battery; are the electricity purchase price and electricity sales price in period t respectively; F op The Z in the equation represents the type of equipment; i is the unit output maintenance parameter of the i-th device; P t i is the output of the ith device in period t; E bat,max is the configured battery energy storage capacity; g d (t) is a segmented 0-1 variable of the discharge depth, and its value of 1 indicates that the discharge depth of the energy storage battery at time t is in the dth segment interval; P buy Indicates the purchased power; P sell represents the power sold; T represents the total dispatching time; B s (N day ) is a binary state variable representing the number of daily cycles; (2) The constraints of the objective function include: The power balance constraint of the household microgrid is as follows: In formula (19): P pv (t), P buy (t), P sell (t) are the photovoltaic output, power purchase and power sales at time t; P in (t), P out (t) are the charging power and discharging power of the energy storage battery at time t respectively; They are the power of the i-th rigid load, the power of the i-th movable load, the power of the i-th interruptible load, and the power of the i-th curtailable load at time t respectively; N1, N2, N3, and N4 refer to the number of rigid loads, the number of translatable loads, the number of interruptible loads, and the number of reducible loads, respectively; Respectively refer to the purchase and sale limits of electricity; buy , δ sell They refer to the 0-1 variables when purchasing and selling electricity, which constrain the microgrid not to sell and purchase electricity at the same time.

10. The method for optimizing energy dispatch of a household microgrid based on time-of-use electricity price and considering demand response according to claim 9, characterized in that: The solution of the microgrid energy optimization scheduling model includes the following steps: Step 1: Data preparation: Collect household load data, historical photovoltaic output data, time-of-use electricity price information and battery parameters; Step 2: Scenario generation: Use Latin hypercube sampling to generate a large number of photovoltaic output scenarios; Apply Kantorovich scenario reduction method to screen out representative typical scenarios based on probability distance; Step 3: Model solution: Use the Yalmip toolkit to build an optimization model in MATLAB, call the Cplex solver to solve the mixed integer linear programming, and output the optimal scheduling strategy for each time period, including battery charging and discharging plan, flexible load start and stop time, and power grid purchase / sale decision; Step 4: Simulation verification and effect evaluation: Input actual household load and photovoltaic output data and run the optimization model.

Images