Distribution network expansion planning method with hydrogen-heat storage in high-proportion photovoltaic scenario

Abstract

The present invention discloses a method for planning the expansion of a distribution network containing hydrogen-heat storage in a high-proportion photovoltaic scenario. In order to effectively characterize the uncertainty of photovoltaic power generation and multi-load energy consumption, a two-layer coupled meteorological clustering model based on the division of heating seasons was established, and an irradiation-load-temperature multi-grid joint scenario generation model was constructed based on each meteorological cluster. In order to solve the problem of insufficient accuracy of traditional photovoltaic output models in high-proportion photovoltaic power generation scenarios and insufficient power demand in high PV output scenarios, which leads to difficulty in accommodating photovoltaic output, a refined distributed photovoltaic system physical model chain was constructed. Considering that the multi-energy coupling of distributed hydrogen-heat storage improves energy utilization efficiency, and balances the contradictions among the economic efficiency of the system's typical scenario operation, the power supply reliability of low-probability high-load scenarios, and the photovoltaic absorption of high PV output scenarios, the distribution network expansion planning results can well solve the economic, reliability and environmental protection problems brought about by high-proportion photovoltaic scenarios on the planning and operation of distribution networks.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network expansion planning, and specifically relates to a distribution network expansion planning method containing hydrogen-heat storage in a high-proportion photovoltaic scenario. Background Art

[0002] Under the "dual carbon" goal, China's new energy penetration rate continues to increase. The randomness and volatility brought about by the high proportion of renewable energy output have led to increased source-load uncertainty. How to accurately characterize the source-load uncertainty scenario and build an economical, reliable, and low-carbon planning model to meet the energy demand of multi-energy coupling is a key issue that needs to be urgently solved in planning the equipment capacity of distributed hydrogen-heat storage systems and expanding lines and transformer capacity under the existing distribution network.

[0003] Existing scenario generation methods can be divided into probabilistic models, classical scenarios, and deep learning generation methods. Based on statistical experience or probability distribution, the probabilistic model method combines the Markov chain sampling method to generate wind, solar, or load scenarios. In order to reduce or optimize large-scale historical scenarios and generate a set of classical scenarios representing the entire area to be solved, the classical scenario generation method uses both data mining and scenario reduction techniques. The deep learning generation method is based on a deep learning framework and can perform deep mining of data to analyze the statistical laws within the data and achieve unsupervised generation of scenarios. However, existing scenario generation methods do not consider how to characterize the uncertainty of multi-node distributed photovoltaic power generation and the uncertainty of energy consumption of multi-node and multi-load in the distribution network.

[0004] Efficient and accurate photovoltaic power output modeling plays a vital role in the planning and operation of distribution networks in high-PV penetration scenarios. Current distribution network planning methods demonstrate good results in specific scenarios. However, in scenarios with high-proportion distributed rooftop PV, PV output is negatively impacted by numerous factors. Meteorological factors such as rainy weather, PV cell temperature, snow cover, and pollution can cause actual PV output to fall below that of traditional PV models, introducing significant randomness and uncertainty. To address these issues, it is necessary to consider the impact of multiple scenarios and meteorological factors on PV power output, and a more refined irradiation-to-PV output model is urgently needed.

[0005] Due to the volatility and uncertainty of distributed photovoltaic power output, when a high proportion of distributed photovoltaic systems are connected to the distribution network, issues such as voltage overshoot, operational stability, and photovoltaic absorption become increasingly prominent. Influenced by the randomness of multiple meteorological factors, comprehensive energy planning models in typical scenarios struggle to support the diverse energy demands and power supply reliability requirements of the distribution network under abnormal weather conditions. However, existing distribution network planning research has not fully considered the power supply reliability issues of distribution networks in scenarios with low probability of high energy loads. Summary of the Invention

[0006] The purpose of the present invention is to provide a distribution network expansion planning method containing hydrogen-heat storage under high-proportion photovoltaic scenarios. First, in order to effectively characterize the uncertainty of photovoltaic power generation and multi-load energy consumption, a two-layer coupled meteorological clustering model based on the division of heating seasons was established, and an irradiation-load-temperature multi-grid joint scenario generation model was constructed based on each meteorological cluster. Then, in order to solve the problem of insufficient accuracy of traditional photovoltaic output models under high-proportion photovoltaic power generation scenarios and insufficient power demand under high PV output scenarios, which leads to difficulty in accommodating photovoltaic output, a refined distributed photovoltaic system physical model chain was constructed. Finally, considering that the distributed hydrogen-heat storage multi-energy coupling improves energy utilization efficiency, and balances the contradictions among the economic operation of the system in typical scenarios, the power supply reliability of low-probability high-load scenarios, and the photovoltaic absorption of high PV output scenarios, the distribution network expansion planning results obtained can well solve the economic, reliability and environmental protection problems brought by high-proportion photovoltaic scenarios to the planning and operation of distribution networks.

[0007] The technical solution adopted by the present invention is: a distribution network expansion planning method for a hydrogen-heat storage system in a high-proportion photovoltaic scenario, which is specifically implemented according to the following steps:

[0008] Step 1: Generate a multi-source load-multi-meteorological-multi-grid joint scenario based on historical meteorological, irradiation, temperature, and load data;

[0009] Step 2: Based on the incident angle model, array plane irradiance conversion model, PVsystem photovoltaic cell temperature model, PVWatts loss model, and photovoltaic system model, a distributed photovoltaic system physical model chain considering multi-dimensional power loss is constructed. The irradiation data of each scenario set generated by the multi-source-load-multi-meteorological-multi-grid joint scenario is converted into photovoltaic output data.

[0010] Step 3: Based on the energy coupling mathematical model of electric heat pump, heat storage tank, alkaline electrolyzer, hydrogen storage tank, solid oxide fuel cell, electric refrigerator, and lithium bromide refrigerator, a multi-energy coupling model of the distributed hydrogen-heat storage system is constructed;

[0011] Step 4: Introduce time-series voltage sensitivity to optimize the site selection of the multi-energy coupling model of the distributed hydrogen-heat storage system. Based on the economic efficiency of distribution network planning under high-proportion photovoltaic scenarios, the operational reliability of low-probability high-load scenarios, and the curtailment parameters of high PV output scenarios, a multi-objective two-layer expansion planning model for the distribution network is constructed. The multi-source-load-multi-meteorological-multi-grid joint scenario and photovoltaic output data are input into the multi-objective two-layer expansion planning model to obtain the distribution network expansion planning results.

[0012] The present invention is also characterized in that:

[0013] The specific process of step 1 is: input historical meteorological information, irradiation, temperature, and load into the ACWGAN-GP model to build a multi-source, multi-meteorological, and multi-node joint scenario generation model.

[0014] The loss functions of the generator and discriminator in ACWGAN-GP are:

[0015] (1)

[0016] (2)

[0017] Where, = 10, = 1; c x Is with data x Related ground truth labels; data x Includes both real and generated data; P r and P G represent the probability distributions of real and generated data, respectively.

[0018] There is a game value function between the generator and the discriminator in ACWGAN-GP V ( G , D ), expressed as:

[0019] (3)

[0020] The specific process of step 2 is:

[0021] The angle of incidence (AOI) refers to the solar angle of incidence, which is defined as the angle between the beam irradiance and the normal to the photovoltaic array surface:

[0022] (4)

[0023] Where, γ is the solar azimuth; Z is the solar zenith angle; γ s is the azimuth angle of the photovoltaic panel surface; β s is the tilt angle of the photovoltaic panel surface;

[0024] Nominal POA irradiance I It is the sum of the beam POA irradiance, the sky diffuse POA irradiance and the ground reflected POA irradiance, expressed as:

[0025] (5)

[0026] Where, I b is the beam POA irradiance; I d is the sky diffuse POA irradiance;I r is the ground reflected POA irradiance;

[0027] Beam POA irradiance I b It is the solar energy that reaches the surface of the photovoltaic array in a straight line from the sun:

[0028] (6)

[0029] Where, E b is the beam irradiance;

[0030] Sky diffuse POA irradiance I d is the solar energy that has been scattered by molecules and particles in the Earth's atmosphere before reaching the subarray surface and is expressed as:

[0031] (7)

[0032] Where, E d is the diffuse irradiance;

[0033] Ground reflected POA irradiance I r It is the solar energy that reaches the array surface after being reflected from the ground. The ground reflected irradiance is the diffuse irradiance, which is a function of the beam normal irradiance and the solar zenith angle, the sky diffuse irradiance and the ground reflectivity, and is expressed as:

[0034] (8)

[0035] Where, is the albedo;

[0036] Taking into account the empirical heat loss factor, the photovoltaic cell temperature and photovoltaic output can be modeled as:

[0037] (9)

[0038] Where, T cell is the photovoltaic cell module temperature; is the absorption coefficient; I is the total incident irradiance (W / m 2 ); T a is the ambient dry bulb temperature ( ℃ ); WS is the wind speed measured at the same height under the determination of wind damage coefficient (m / s); U c is the comprehensive heat loss factor coefficient; Uv is the comprehensive heat loss coefficient affected by wind; is the module external efficiency;

[0039] The total loss of the photovoltaic system is not the sum of the individual losses, which is obtained by multiplying each loss by L i (%) is calculated based on the reduction caused by the following formula:

[0040] (10)

[0041] Where, L i Expressed as i The percentage of reduction in system power output caused by this loss (%);

[0042] The PVWatts DC power model is expressed as:

[0043] (11)

[0044] Where, I is the irradiance transmitted to the photovoltaic cell (W / m 2 ); P dc0 For photovoltaic cell modules at 1000W / m 2 and power at reference temperature (W); γ pdc is the power temperature coefficient, T ref is the battery reference temperature;

[0045] The PVWatts inverter model is expressed as:

[0046] (12)

[0047] in, (13)

[0048] (14)

[0049] (15)

[0050] Where, P dc The DC output power of the photovoltaic cell; P dcl0 is the DC input limit of the inverter; is the nominal inverter efficiency; For reference inverter efficiency, PVWatts defines it as 0.9637; P ac0 is the rated output AC power of the inverter;

[0051] Formulas (4)-(15) are used as a physical model chain of distributed photovoltaic systems considering multivariate power losses;

[0052] The multi-source-load-multi-weather-multi-grid joint scenario is input into the distributed photovoltaic system physical model chain considering multiple power losses to obtain the irradiation data of each scenario set and convert it into photovoltaic output data.

[0053] The specific process of step 3 is as follows:

[0054] The multi-energy input and output balance relationship of the distributed hydrogen-heat storage system is expressed as:

[0055] (16)

[0056] Where, L el 、 L cl 、 L hl They are electricity, cooling and heating loads respectively; P grid Transmitting power to the grid; P PV is the PV output power; P SOFC The electrical power output of the solid oxide fuel cell; P EHP 、 P AE 、 P ER are the electricity loads of the electric heat pump, alkaline electrolyzer, and electric refrigerator respectively; C ER 、 C LBR are the cooling power output by the electric refrigerator and lithium bromide refrigerator respectively; H HST,out is the thermal power output of the heat storage tank;

[0057] The mathematical model of each energy coupling device in the distributed hydrogen-heat storage system is:

[0058] (17)

[0059] Where, H EHP 、 H SOFC are the thermal powers output by the electric heat pump and solid oxide fuel cell respectively; H TST,in is the thermal power absorbed by the heat storage tank; M AE For alkaline electrolysis cells The mass of hydrogen produced in a given time; M HST For hydrogen storage tanks The quality of hydrogen output within a certain time period; L is the lower calorific value of hydrogen; H SOFC is the thermal power output of the solid oxide fuel cell; 、 are the output heat-to-electricity ratio and heat recovery coefficient of the solid oxide fuel cell respectively; Energy conversion efficiency of the equipment; is the heat transfer coefficient.

[0060] The multi-objective two-layer extended planning model of the distribution network in step 4 includes an upper-layer planning model, a lower-layer planning model and constraints;

[0061] The upper-level planning model uses the capacity of the planned equipment, including electric heat pumps, heat storage tanks, alkaline electrolyzers, hydrogen storage tanks, solid oxide fuel cells, electric refrigerators, and lithium bromide refrigerators, in the multi-energy coupling model of each distributed hydrogen-heat storage system as decision variables, and minimizes the annual comprehensive cost of the system as the objective function, which is expressed as:

[0062] (18)

[0063] Where, C eco is the annual economic cost of the system; C re The cost of power supply reliability; C apv Penalty cost for abandoned light;

[0064] The lower-level planning model takes the optimized output of each device in the distributed hydrogen-heat storage system multi-energy coupling model as the decision variable and the minimum daily operating cost of each typical scenario in the distributed hydrogen-heat storage system multi-energy coupling model as the objective function, which can be expressed as:

[0065] (31)

[0066] (32)

[0067] Where, For the scene s Down t The power purchased by the upper power grid at any given moment; for t The electricity purchase price at any given moment; is the time interval;

[0068] (33)

[0069] Where, 、 、 They are time-of-use electricity price, heating price and cooling price; For the scene s Down t Solid oxide fuel cell output at all times; For the scene s Down t The heat storage tank supplies heat at all times; For the scene s Down t Cooling load at all times.

[0070] In the upper-level planning model, there are:

[0071] 1) The annual economic cost of the system is expressed as:

[0072] (19)

[0073] Where, C inv is the annualized investment cost of the system; C main is the total annual maintenance cost of the equipment; C ope is the annual operating cost of the system;

[0074] (20)

[0075] Where, C inv is the annualized investment cost during the planning period; is the investment cost per unit capacity of the equipment; The equipment investment amount is set to an integer; Invest in rated power for a single device; is the investment cost per unit length of the line; Invest 0-1 variables for the line; For the j The length of the line;

[0076] The annual investment cost calculation formula considering the equipment capital recovery factor CRF is:

[0077] (twenty one)

[0078] Where, r is the interest rate; L The service life of the equipment;

[0079] The annual equipment maintenance cost of the system considers the annual maintenance cost of each device and transformer in the multi-energy coupling model of the distributed hydrogen-heat storage system, namely:

[0080] (twenty two)

[0081] Where, The annual maintenance cost per unit capacity of the equipment;

[0082] The annual operating cost of the system is the sum of the daily operating costs of the distributed hydrogen-heat storage system multi-energy coupling model in multiple scenarios within a year;

[0083] (twenty three)

[0084] Where, The number of days for each typical scenario; For the scene s The daily electricity purchase cost of the multi-energy coupling model of the distributed hydrogen-heat storage system; For the scene s Daily energy sales revenue of the multi-energy coupling model of the distributed hydrogen-heat storage system;

[0085] 2) Power supply reliability cost in low-probability high-load scenarios

[0086] The system power restriction index can comprehensively reflect the power outage time and load shortage of each node in the distribution network. The system power restriction index can be converted into power restriction cost as the annual power supply reliability cost of the distribution network to establish the power supply reliability target. C re Expressed as:

[0087] (twenty four)

[0088] (25)

[0089] Where, c p Penalty cost for unit power shortage; E lack The annual power deficit of the system; For the i node t The power shortage at the moment; For the i The existing transformer capacity of the node; For the i Rated capacity of the transformer to be expanded at the node;

[0090] When the line flow is greater than the line capacity, the line power supply reliability is described by considering the overload power of the distribution network line flow in a low-probability high-load scenario, that is:

[0091] (26)

[0092] (27)

[0093] Where, For the l Overload power of the line; For the l Lines in a low probability high load scenario t Moment power; For the l Existing line capacity of lines; For the l The rated capacity of the line to be expanded;

[0094] 3) Penalty cost for curtailment in high PV output scenarios

[0095] Introducing a penalty factor for abandoned solar power and establishing an objective function for abandoned solar power cost in high PV output scenarios C apv , expressed as:

[0096] (28)

[0097] (29)

[0098] (30)

[0099] Where, c a is the penalty cost per unit of curtailed solar power; The abandoned solar power in high PV output scenarios is caused by insufficient regulation capability of the multi-energy coupling model of the distributed hydrogen-heat storage system; This refers to the abandoned optical power caused by insufficient line capacity in high PV output scenarios; When the system is powered only by photovoltaic power l Active power flowing from the PV node to the load node.

[0100] The constraints are:

[0101] 1) Equipment capacity constraints of the multi-energy coupling model of distributed hydrogen-heat storage system

[0102] The rated capacity of the multi-energy coupling model equipment of the planned distributed hydrogen-heat storage system should meet the following constraints:

[0103] (34)

[0104] Where, Set a capacity cap for each piece of equipment investment;

[0105] 2) Energy balance constraints

[0106] The multi-energy coupling model of the distributed hydrogen-heat storage system should satisfy the multi-energy supply and demand balance of each node, so the planning model should consider the energy balance constraint shown in formula (17);

[0107] 3) Hydrogen storage tank energy constraints

[0108] The hydrogen storage tank should meet the hydrogen mass balance requirements:

[0109] (35)

[0110] Where, 、 、 、 They are t The quality of residual gas, intake gas, discharge gas and gas loss in hydrogen storage tanks at all times;

[0111] exist t The hydrogen storage mass at any moment should be less than or equal to the rated capacity of the hydrogen storage tank;

[0112] (36)

[0113] Where, is the rated capacity of the hydrogen storage tank;

[0114] exist t The filling and discharging rate of the hydrogen storage tank should be less than or equal to the rated filling and discharging rate of the hydrogen storage tank at all times;

[0115] (37)

[0116] (38)

[0117] Where, is the rated capacity of the hydrogen storage tank;

[0118] 4) Thermal storage tank energy constraints

[0119] The heat storage tank should satisfy the heat balance shown in formula (40) and t The heat storage capacity at any moment should be less than or equal to the rated capacity of the heat storage tank;

[0120] (39)

[0121] (40)

[0122] Where, Q t 、 、 、 They are tResidual heat, input heat, output heat and heat loss of the thermal storage tank at all times; is the rated capacity of the heat storage tank;

[0123] 5) Solid oxide fuel cell output power constraints

[0124] Fuel cells in t The electrical power output at any time should be less than or equal to the rated capacity of the fuel cell, expressed as:

[0125] (41)

[0126] Where, P SOFC,e is the rated capacity of the solid oxide fuel cell;

[0127] 6) Transmission line capacity constraints

[0128] (42)

[0129] Where, P ij,max 、 Q ij,max for ij Maximum active and reactive power of lines between nodes; S e,ij for ij Rated capacity of inter-node lines;

[0130] 7) Node voltage constraints

[0131] The voltage of each node in the distribution network should meet the following constraints:

[0132] (43)

[0133] Where, U i For the distribution network i Node voltage; U i,min 、 U i,max Respectively i Node voltage lower and upper limits.

[0134] The beneficial effects of the present invention are:

[0135] (1) The proposed multi-source load-multi-meteorological-multi-grid joint scenario generation method helps to analyze the actual probability distribution of daily operation of the distribution network, and the generated scenarios are more consistent with historical scenarios.

[0136] (2) A physical model chain of distributed photovoltaic systems that takes into account multiple power losses is adopted, and the impact of irradiation, temperature, wind speed and multiple power losses on photovoltaic power output under actual photovoltaic operation scenarios is considered to meet the planning needs of hydrogen-heat storage multi-energy coupling distribution network under high photovoltaic power penetration scenarios.

[0137] (3) The proposed distributed hydrogen-heat storage multi-energy coupling distribution network multi-objective two-layer expansion planning model takes into account that the distributed hydrogen-heat storage multi-energy coupling improves energy utilization efficiency and balances the conflicts among the system's typical scenario operation economy, the power supply reliability of low-probability high-load scenarios, and the photovoltaic absorption of high PV output scenarios. The distribution network expansion planning results obtained can well solve the economic, reliability and environmental protection problems brought about by high-proportion photovoltaic scenarios on distribution network planning and operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0138] Figure 1 This is a flow chart of the distribution network expansion planning method for hydrogen-heat storage in a high-proportion photovoltaic scenario of the present invention.

[0139] Figure 2 It is a model chain structure diagram;

[0140] Figure 3 It is the DHTSS structure and input and output diagram;

[0141] Figure 4 This is a PDF comparison chart of scene generation results using different methods. DETAILED DESCRIPTION

[0142] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0143] The invention provides a method for planning the expansion of a distribution network with hydrogen-heat storage in a high-proportion photovoltaic scenario. Figure 1 As shown, please follow the steps below:

[0144] Step 1: Generate a multi-source load, multi-meteorological, and multi-grid joint scenario based on historical meteorological, irradiation, temperature, and load data. The specific process is as follows: input historical meteorological, irradiation, temperature, and load data into the ACWGAN-GP model to construct a multi-source load, multi-meteorological, and multi-node joint scenario generation model.

[0145] The loss functions of the generator and discriminator in ACWGAN-GP are:

[0146] (1)

[0147] (2)

[0148] Where, = 10, = 1;c x Is with data x Related ground truth labels; data x Includes both real and generated data; P r and P G represent the probability distributions of real and generated data, respectively.

[0149] ACWGAN-GP constructs a game value function V ( G , D ) to build the generator G and the discriminator D A game model between the generator and the discriminator so that they can be trained simultaneously. V ( G , D ), expressed as:

[0150] (3).

[0151] Step 2: Based on the incident angle model, array plane irradiance conversion model, PV system photovoltaic cell temperature model, PVWatts loss model and photovoltaic system model, a distributed photovoltaic system physical model chain considering multivariate power loss is constructed. The structure is as follows: Figure 2 As shown in the figure, the irradiation data of each scenario set generated by the multi-source load-multi-meteorological-multi-grid joint scenario is converted into photovoltaic output data; the specific process is:

[0152] The angle of incidence (AOI) refers to the solar angle of incidence, which is defined as the angle between the beam irradiance and the normal to the photovoltaic array surface:

[0153] (4)

[0154] Where, γ is the solar azimuth; Z is the solar zenith angle; γ s is the azimuth angle of the photovoltaic panel surface; β s is the tilt angle of the photovoltaic panel surface;

[0155] Nominal POA irradiance I It is the sum of the beam POA irradiance, the sky diffuse POA irradiance and the ground reflected POA irradiance, expressed as:

[0156] (5)

[0157] Where, I bis the beam POA irradiance; I d is the sky diffuse POA irradiance; I r is the ground reflected POA irradiance;

[0158] Beam POA irradiance I b It is the solar energy that reaches the surface of the photovoltaic array in a straight line from the sun:

[0159] (6)

[0160] Where, E b is the beam irradiance;

[0161] Sky diffuse POA irradiance I d is the solar energy that has been scattered by molecules and particles in the Earth's atmosphere before reaching the subarray surface and is expressed as:

[0162] (7)

[0163] Where, E d is the diffuse irradiance;

[0164] Ground reflected POA irradiance I r It is the solar energy that reaches the array surface after being reflected from the ground. The ground reflected irradiance is the diffuse irradiance, which is a function of the beam normal irradiance and the solar zenith angle, the sky diffuse irradiance and the ground reflectivity, and is expressed as:

[0165] (8)

[0166] Where, is the albedo;

[0167] Taking into account the empirical heat loss factor, the photovoltaic cell temperature and photovoltaic output can be modeled as:

[0168] (9)

[0169] Where, T cell is the photovoltaic cell module temperature; is the absorption coefficient; I is the total incident irradiance (W / m 2 ); T a is the ambient dry bulb temperature ( ℃ ); WS is the wind speed measured at the same height under the determination of wind damage coefficient (m / s);U c is the comprehensive heat loss factor coefficient; U v is the comprehensive heat loss coefficient affected by wind; is the module external efficiency;

[0170] The total loss of the photovoltaic system is not the sum of the individual losses, which is obtained by multiplying each loss by L i The default total system loss is set to 14%, and the calculation formula is:

[0171] (10)

[0172] Where, L i Expressed as i The percentage of reduction in system power output caused by this loss (%);

[0173] The PVWatts DC power model is expressed as:

[0174] (11)

[0175] Where, I is the irradiance transmitted to the photovoltaic cell (W / m 2 ); P dc0 For photovoltaic cell modules at 1000W / m 2 and power at reference temperature (W); γ pdc is the power temperature coefficient, usually -0.002 to -0.005 per degree Celsius (1 / °C); T ref is the battery reference temperature, which PVWatts defines as 25°C.

[0176] The PVWatts inverter model is expressed as:

[0177] (12)

[0178] in, (13)

[0179] (14)

[0180] (15)

[0181] Where, P dc The DC output power of the photovoltaic cell; P dcl0is the DC input limit of the inverter; is the nominal inverter efficiency, which is 0.96 by default; For reference inverter efficiency, PVWatts defines it as 0.9637; P ac0 is the rated output AC power of the inverter;

[0182] Formulas (4)-(15) are used as a physical model chain of distributed photovoltaic systems considering multivariate power losses;

[0183] The multi-source-load-multi-weather-multi-grid joint scenario is input into the distributed photovoltaic system physical model chain considering multiple power losses to obtain the irradiation data of each scenario set and convert it into photovoltaic output data.

[0184] Step 3: Based on the energy coupling mathematical model of electric heat pump, heat storage tank, alkaline electrolyzer, hydrogen storage tank, solid oxide fuel cell, electric refrigerator, lithium bromide refrigerator, such as Figure 3 As shown in the figure, a multi-energy coupling model of a distributed hydrogen-heat storage system is constructed; the specific process is as follows:

[0185] The multi-energy input and output balance relationship of the distributed hydrogen-heat storage system is expressed as:

[0186] (16)

[0187] Where, L el 、 L cl 、 L hl They are electricity, cooling and heating loads respectively; P grid Transmitting power to the grid; P PV is the PV output power; P SOFC The electrical power output of the solid oxide fuel cell; P EHP 、 P AE 、 P ER are the electricity loads of the electric heat pump, alkaline electrolyzer, and electric refrigerator respectively; C ER 、 C LBR are the cooling power output by the electric refrigerator and lithium bromide refrigerator respectively; H HST,out is the thermal power output of the heat storage tank;

[0188] The mathematical model of each energy coupling device in the distributed hydrogen-heat storage system is:

[0189] (17)

[0190] Where, H EHP 、 H SOFC are the thermal powers output by the electric heat pump and solid oxide fuel cell respectively; H TST,in is the thermal power absorbed by the heat storage tank; M AE For alkaline electrolysis cells The mass of hydrogen produced in a given time; M HST For hydrogen storage tanks The quality of hydrogen output within a certain time period; L is the lower calorific value of hydrogen; H SOFC is the thermal power output of the solid oxide fuel cell; 、 are the output heat-to-electricity ratio and heat recovery coefficient of the solid oxide fuel cell respectively; Energy conversion efficiency of the equipment; is the heat transfer coefficient.

[0191] Step 4: Introduce time-series voltage sensitivity to optimize the site selection of the multi-energy coupling model of the distributed hydrogen-heat storage system. Based on the economic efficiency of distribution network planning under high-proportion photovoltaic scenarios, the operational reliability of low-probability high-load scenarios, and the curtailment parameters of high PV output scenarios, a multi-objective two-layer expansion planning model for the distribution network is constructed. The multi-source-load-multi-meteorological-multi-grid joint scenario and photovoltaic output data are input into the multi-objective two-layer expansion planning model to obtain the distribution network expansion planning results.

[0192] The multi-objective two-layer extended planning model for distribution network includes upper-layer planning model, lower-layer planning model and constraints;

[0193] Upper-level planning model

[0194] Aiming at the economic efficiency of distribution network planning under high-proportion photovoltaic scenarios, the operational reliability of low-probability high-load scenarios, and the problem of curtailment in high PV output scenarios, an upper-level planning model is constructed. The capacity of the planned equipment in the multi-energy coupling model of each distributed hydrogen-heat storage system, including electric heat pumps, thermal storage tanks, alkaline electrolyzers, hydrogen storage tanks, solid oxide fuel cells, electric refrigerators, and lithium bromide refrigerators, is used as the decision variable, and the objective function is to minimize the annual comprehensive cost of the system, which is expressed as:

[0195] (18)

[0196] Where, C eco is the annual economic cost of the system;C re The cost of power supply reliability; C apv Penalty cost for abandoned light;

[0197] Lower-level planning model

[0198] The lower-level planning model takes the optimized output of each device in the distributed hydrogen-heat storage system multi-energy coupling model as the decision variable and the minimum daily operating cost of each typical scenario in the distributed hydrogen-heat storage system multi-energy coupling model as the objective function, which can be expressed as:

[0199] (31)

[0200] (32)

[0201] Where, For the scene s Down t The power purchased by the upper power grid at any given moment; for t The electricity purchase price at any given moment; is the time interval;

[0202] (33)

[0203] Where, 、 、 They are time-of-use electricity price, heating price and cooling price; For the scene s Down t Solid oxide fuel cell output at all times; For the scene s Down t The heat storage tank supplies heat at all times; For the scene s Down t Cooling load at all times.

[0204] In the upper-level planning model, there are:

[0205] 1) The annual economic cost of the system is expressed as:

[0206] (19)

[0207] Where, C inv is the annualized investment cost of the system; C main is the total annual maintenance cost of the equipment; C ope is the annual operating cost of the system;

[0208] (20)

[0209] Where, C inv is the annualized investment cost during the planning period; is the investment cost per unit capacity of the equipment; The equipment investment amount is set to an integer; Invest in rated power for a single device; is the investment cost per unit length of the line; Invest 0-1 variables for the line; For the j The length of the line;

[0210] The annual investment cost calculation formula considering the equipment capital recovery factor CRF is:

[0211] (twenty one)

[0212] Where, r is the interest rate; L The service life of the equipment;

[0213] The annual equipment maintenance cost of the system considers the annual maintenance cost of each device and transformer in the multi-energy coupling model of the distributed hydrogen-heat storage system, namely:

[0214] (twenty two)

[0215] Where, The annual maintenance cost per unit capacity of the equipment;

[0216] The annual operating cost of the system is the sum of the daily operating costs of the distributed hydrogen-heat storage system multi-energy coupling model in multiple scenarios within a year;

[0217] (twenty three)

[0218] Where, The number of days for each typical scenario; For the scene s The daily electricity purchase cost of the multi-energy coupling model of the distributed hydrogen-heat storage system; For the scene s Daily energy sales revenue of the multi-energy coupling model of the distributed hydrogen-heat storage system;

[0219] 2) Power supply reliability cost in low-probability high-load scenarios

[0220] The system power restriction index can comprehensively reflect the power outage time and load shortage of each node in the distribution network. The system power restriction index can be converted into power restriction cost as the annual power supply reliability cost of the distribution network to establish the power supply reliability target. C reExpressed as:

[0221] (twenty four)

[0222] (25)

[0223] Where, c p Penalty cost for unit power shortage; E lack The annual power deficit of the system; For the i node t The power shortage at the moment; For the i The existing transformer capacity of the node; For the i Rated capacity of the transformer to be expanded at the node;

[0224] When the line flow is greater than the line capacity, the line power supply reliability is described by considering the overload power of the distribution network line flow in a low-probability high-load scenario, that is:

[0225] (26)

[0226] (27)

[0227] Where, For the l Overload power of the line; For the l Lines in a low probability high load scenario t Moment power; For the l Existing line capacity of lines; For the l The rated capacity of the line to be expanded;

[0228] 3) Penalty cost for curtailment in high PV output scenarios

[0229] Introducing a penalty factor for abandoned solar power and establishing an objective function for abandoned solar power cost in high PV output scenarios C apv , expressed as:

[0230] (28)

[0231] (29)

[0232] (30)

[0233] Where, c ais the penalty cost per unit of curtailed solar power; The abandoned solar power in high PV output scenarios is caused by insufficient regulation capability of the multi-energy coupling model of the distributed hydrogen-heat storage system; This refers to the abandoned optical power caused by insufficient line capacity in high PV output scenarios; When the system is powered only by photovoltaic power l Active power flowing from the PV node to the load node.

[0234] The constraints are:

[0235] 1) Equipment capacity constraints of the multi-energy coupling model of distributed hydrogen-heat storage system

[0236] The rated capacity of the multi-energy coupling model equipment of the planned distributed hydrogen-heat storage system should meet the following constraints:

[0237] (34)

[0238] Where, Set a capacity cap for each piece of equipment investment;

[0239] 2) Energy balance constraints

[0240] The multi-energy coupling model of the distributed hydrogen-heat storage system should satisfy the multi-energy supply and demand balance of each node, so the planning model should consider the energy balance constraint shown in formula (17);

[0241] 3) Hydrogen storage tank energy constraints

[0242] The hydrogen storage tank should meet the hydrogen mass balance requirements:

[0243] (35)

[0244] Where, 、 、 、 They are t The quality of residual gas, intake gas, discharge gas and gas loss in hydrogen storage tanks at all times;

[0245] exist t The hydrogen storage mass at any moment should be less than or equal to the rated capacity of the hydrogen storage tank;

[0246] (36)

[0247] Where, is the rated capacity of the hydrogen storage tank;

[0248] exist t The filling and discharging rate of the hydrogen storage tank should be less than or equal to the rated filling and discharging rate of the hydrogen storage tank at all times;

[0249] (37)

[0250] (38)

[0251] Where, is the rated capacity of the hydrogen storage tank;

[0252] 4) Thermal storage tank energy constraints

[0253] The heat storage tank should satisfy the heat balance shown in formula (40) and t The heat storage capacity at any moment should be less than or equal to the rated capacity of the heat storage tank;

[0254] (39)

[0255] (40)

[0256] Where, Q t 、 、 、 They are t Residual heat, input heat, output heat and heat loss of the thermal storage tank at all times; is the rated capacity of the heat storage tank;

[0257] 5) Solid oxide fuel cell output power constraints

[0258] Fuel cells in t The electrical power output at any time should be less than or equal to the rated capacity of the fuel cell, expressed as:

[0259] (41)

[0260] Where, P SOFC,e is the rated capacity of the solid oxide fuel cell;

[0261] 6) Transmission line capacity constraints

[0262] (42)

[0263] Where, P ij,max 、 Q ij,max for ij Maximum active and reactive power of lines between nodes; S e,ij for ij Rated capacity of inter-node lines;

[0264] 7) Node voltage constraints

[0265] The voltage of each node in the distribution network should meet the following constraints:

[0266] (43)

[0267] Where, U i For the distribution network i Node voltage; U i,min 、 U i,max Respectively i Node voltage lower and upper limits.

[0268] 1) Multi-source load, multi-weather, and multi-grid joint scenario generation

[0269] Figure 4 The effectiveness of the ACWGAN-GP method was demonstrated by comparing the random feature PDFs of scenes generated by ACWGAN-GP with those of raw data (RD) and a modular denoising variational autoencoder (MDVAE). The results showed that the scenes generated by the proposed method had a higher similarity in probability density function (PDF) than the RD method compared to the MDVAE method.

[0270] 2) Multi-objective planning results of distribution network planning model

[0271] In order to verify the applicability and superiority of the multi-objective two-layer expansion planning model of the hydrogen-heat storage system proposed in this invention in the scenario of high-proportion distributed photovoltaic power access, five cases are set for comparative analysis: (1) Distribution network expansion planning that does not consider the generation of source-load uncertainty scenarios and the coupling of cold, heat and electricity, and configures distributed energy storage systems at the photovoltaic installation location according to the lower limit of policy requirements (10% of photovoltaic installed capacity); (2) Distribution network expansion planning that does not consider the generation of source-load uncertainty scenarios and the coupling of cold, heat and electricity, and takes into account the optimal configuration of distributed battery energy storage capacity; (3) Distribution network expansion planning that considers the optimal configuration of the hydrogen-heat storage system but does not consider the reliability target of the low-probability high-load scenario; (4) Distribution network expansion planning that considers the optimal configuration of the hydrogen-heat storage system but does not consider the photovoltaic absorption target; (5) Distribution network expansion planning that considers the optimal configuration of the hydrogen-heat storage system and considers multiple objectives.

[0272] Case 1 and Case 2 performed deterministic planning calculations for the distribution network under the maximum source-load scenario in 2030. The cost of distributed battery energy storage in Case 2 is much higher than in Case 1, as shown in Table 1.

[0273] Table 1

[0274]

[0275] Cases 3, 4, and 5, based on the results of multi-source, multi-load, multi-weather, and multi-grid scenarios, optimize the energy storage capacity configuration of each DHTSS device and plan the line and transformer capacities. Case 3 plans distribution network lines and transformers under a typical daily scenario. Case 4 builds on Case 3 by considering the distribution network line flow distribution under a low-probability high-load scenario and expanding the line and transformer planning. Case 5 builds on Case 4 by considering the impact of line capacity limitations on PV consumption and expanding the line and transformer planning. The investment economics of Cases 3-5 are shown in Table 2.

[0276] Table 2

[0277]

[0278] 3) Optimized operation results of the hydrogen-heat storage system

[0279] The time-of-use electricity purchase and sale price of the hydrogen-heat storage system is shown in Table 3, and the unit calorific value price is 0.223 yuan / kWh.

[0280] Table 3

[0281]

[0282] Table 4

[0283]

[0284] Based on the evaluation indicators of Cases 1-5 in Table 4-6, a comparative analysis is conducted on the system investment and operation economy in typical scenarios, the power supply reliability in low-probability high-load scenarios, and the system photovoltaic absorption rate in high PV output scenarios.

[0285] ① Comparison of Economic Benefits of Typical Scenarios. As shown in Table 4, the distributed battery energy storage systems in Cases 1 and 2 only absorb the excess PV power output at their respective nodes, resulting in an annual electricity purchase cost of zero. Since Case 1 only configured energy storage capacity at the minimum policy requirement, its ability to absorb excess PV power was limited, resulting in an annual energy sales revenue 79.44% lower than that of Case 2. Compared to Cases 1-2, Cases 3-5, while incurring increased annual electricity purchase costs of 30.4292 million RMB, 30.3169 million RMB, and 30.6647 million RMB, respectively, increased their annual comprehensive returns by 65.5%, 63.05%, and 53.79%, respectively, due to diversified energy sales and lower annualized energy storage system investment costs. Compared to Case 3, Case 4 considers power supply reliability targets for low-probability high-load scenarios, resulting in increased line and transformer investment costs, resulting in lower annual comprehensive returns than Case 3. Compared to Case 4, Case 5 increases the DHTSS equipment configuration capacity and the capacity of the PV-to-load node lines to accommodate high PV output scenarios. Therefore, Case 5 has the highest annual investment cost, but its annual comprehensive benefits are lower than those of Cases 3 and 4. Because Case 5 couples the distribution network with the heat network, it offers greater operational flexibility. Furthermore, the increased capacity of the PV-to-load node lines reduces the long-term planning costs for distribution network expansion.

[0286] Table 5

[0287]

[0288] ② Regarding power supply reliability in low-probability high-load scenarios, Table 5 shows that because Cases 1 and 2 implemented deterministic planning of the distribution network under the 2030 maximum load scenario, there was no power curtailment in low-probability high-load scenarios. However, Case 3, considering only the typical scenario system economics and without upgrading or expanding the upstream power grid's source measurement line capacity, experienced the highest power curtailment in low-probability high-load scenarios. Cases 4 and 5, however, considered reliability objectives and rationally expanded the upstream power grid's source measurement line capacity. Consequently, they demonstrated good operational reliability in low-probability high-load scenarios and demonstrated a stronger ability to mitigate risks to the safe and stable operation of the distribution network.

[0289] Table 6

[0290]

[0291] Regarding PV absorption in high PV output scenarios, Table 6 shows that Case 1 only configures distributed battery energy storage according to the lower limit of policy requirements, resulting in limited PV absorption capacity in typical scenarios. The PV absorption rate in high PV output scenarios is only 50.31%. Case 2 optimizes the configuration of the distributed energy storage system based on typical scenarios. This fully absorbs PV output in typical scenarios, but still results in significant curtailment in high PV output scenarios. Cases 3 and 4 do not consider PV absorption targets in high PV output scenarios. Due to certain line capacity limitations, the PV absorption rates in typical scenarios do not reach 100%. Due to DHTSS capacity limitations, the PV absorption rates in high PV output scenarios are only 79.39% and 80.26%. Case 5 considers PV absorption targets and increases the capacity of various DHTSS devices, resulting in stronger regulation capabilities in high PV output scenarios. Furthermore, the line capacity between PV nodes and load nodes is rationally expanded, resulting in excellent PV absorption.

[0292] Taking all factors into consideration, Case 5 studied in this paper has good results in terms of the system's annual comprehensive economy under typical scenarios, power supply reliability under low-probability high-load scenarios, and photovoltaic absorption under high PV output scenarios.

Claims

1. A distribution network expansion planning method for a hydrogen-heat storage system in a high-proportion photovoltaic scenario, characterized in that: Please follow the steps below to implement it: Step 1: Generate a multi-source load-multi-meteorological-multi-grid joint scenario based on historical meteorological, irradiation, temperature, and load data; Step 2: Based on the incident angle model, array plane irradiance conversion model, PVsystem photovoltaic cell temperature model, PVWatts loss model, and photovoltaic system model, a distributed photovoltaic system physical model chain considering multi-dimensional power loss is constructed. The irradiation data of each scenario set generated by the multi-source-load-multi-meteorological-multi-grid joint scenario is converted into photovoltaic output data. Step 3: Based on the energy coupling mathematical model of electric heat pump, heat storage tank, alkaline electrolyzer, hydrogen storage tank, solid oxide fuel cell, electric refrigerator, and lithium bromide refrigerator, a multi-energy coupling model of the distributed hydrogen-heat storage system is constructed; Step 4: Introduce time-series voltage sensitivity to optimize the site selection of the multi-energy coupling model of the distributed hydrogen-heat storage system. A multi-objective, two-layer expansion planning model for the distribution network is constructed based on the economic efficiency of distribution network planning under high-proportion photovoltaic scenarios, the operational reliability of low-probability high-load scenarios, and the curtailment parameters of high PV output scenarios. The multi-source-load, multi-weather, and multi-grid combined scenarios and photovoltaic output data are input into the multi-objective, two-layer expansion planning model to obtain the distribution network expansion planning results. The specific process of step 1 is as follows: historical meteorological data, irradiation, temperature, and load are input into the ACWGAN-GP model to build a multi-source, multi-load, multi-meteorological, and multi-node joint scenario generation model; The distribution network multi-objective two-layer extended planning model in step 4 includes an upper-layer planning model, a lower-layer planning model and constraints; The upper-level planning model uses the capacity of the electric heat pump, heat storage tank, alkaline electrolyzer, hydrogen storage tank, solid oxide fuel cell, electric refrigerator, and lithium bromide refrigerator to be planned in each distributed hydrogen-heat storage system multi-energy coupling model as decision variables, and minimizes the annual comprehensive cost of the system as the objective function. The lower-level planning model uses the optimized output of each device in the distributed hydrogen-heat storage system multi-energy coupling model as the decision variable, and minimizes the daily operating cost of each typical scenario of the distributed hydrogen-heat storage system multi-energy coupling model as the objective function.

2. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 1 is characterized in that: The loss functions of the generator and discriminator in the ACWGAN-GP are: (1) (2) Where, = 10, = 1; c x Is with data x Related ground-truth labels; data x Includes both real and generated data; P r and P G represent the probability distributions of real and generated data, respectively.

3. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 1 is characterized in that: There is a game value function between the generator and the discriminator in the ACWGAN-GP V ( G , D ), expressed as: (3)。 4. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 1 is characterized in that: The specific process of step 2 is: The angle of incidence (AOI) refers to the solar angle of incidence, which is defined as the angle between the beam irradiance and the normal to the photovoltaic array surface: (4) Where, γ is the solar azimuth; Z is the solar zenith angle; γ s is the azimuth angle of the photovoltaic panel surface; β s is the tilt angle of the photovoltaic panel surface; Nominal POA irradiance I It is the sum of the beam POA irradiance, the sky diffuse POA irradiance and the ground reflected POA irradiance, expressed as: (5) Where, I b is the beam POA irradiance; I d is the sky diffuse POA irradiance; I r is the ground reflected POA irradiance; Beam POA irradiance I b It is the solar energy that reaches the surface of the photovoltaic array in a straight line from the sun: (6) Where, E b is the beam irradiance; Sky diffuse POA irradiance I d is the solar energy that has been scattered by molecules and particles in the Earth's atmosphere before reaching the subarray surface and is expressed as: (7) Where, E d is the diffuse irradiance; Ground reflected POA irradiance I r It is the solar energy that reaches the array surface after being reflected from the ground. The ground reflected irradiance is the diffuse irradiance, which is a function of the beam normal irradiance and the solar zenith angle, the sky diffuse irradiance and the ground reflectivity, and is expressed as: (8) Where, is the albedo; Taking into account the empirical heat loss factor, the photovoltaic cell temperature and photovoltaic output can be modeled as: (9) Where, T cell is the photovoltaic cell module temperature; is the absorption coefficient; I is the total incident irradiance (W / m 2 ); T a is the ambient dry bulb temperature ( ℃ ); WS is the wind speed measured at the same height under the determination of wind damage coefficient (m / s); U c is the comprehensive heat loss factor coefficient; U v is the comprehensive heat loss coefficient affected by wind; is the module external efficiency; The total loss of the photovoltaic system is not the sum of the individual losses, which is obtained by multiplying each loss by L i (%) is calculated based on the reduction caused by the following formula: (10) Where, L i Expressed as i The percentage of reduction in system power output caused by this loss (%); The PVWatts DC power model is expressed as: (11) Where, I is the irradiance transmitted to the photovoltaic cell (W / m 2 ); P dc0 For photovoltaic cell modules at 1000W / m 2 and power at reference temperature (W); γ pdc is the power temperature coefficient, T ref is the battery reference temperature; The PVWatts inverter model is expressed as: (12) in, (13) (14) (15) Where, P dc The DC output power of the photovoltaic cell; P dcl0 is the DC input limit of the inverter; is the nominal inverter efficiency; For reference inverter efficiency, PVWatts defines it as 0.9637; P ac0 is the rated output AC power of the inverter; Formulas (4)-(15) are used as a physical model chain of distributed photovoltaic systems considering multivariate power losses; The multi-source-load-multi-weather-multi-grid joint scenario is input into the distributed photovoltaic system physical model chain considering multiple power losses to obtain the irradiation data of each scenario set and convert it into photovoltaic output data.

5. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 1 is characterized in that: The specific process of step 3 is as follows: The multi-energy input and output balance relationship of the distributed hydrogen-heat storage system is expressed as: (16) Where, L el 、 L cl 、 L hl They are electricity, cooling and heating loads respectively; P grid Transmitting power to the grid; P PV is the PV output power; P SOFC The electrical power output of the solid oxide fuel cell; P EHP 、 P AE 、 P ER are the electricity loads of the electric heat pump, alkaline electrolyzer, and electric refrigerator respectively; C ER 、 C LBR are the cooling power output by the electric refrigerator and lithium bromide refrigerator respectively; H HST,out is the thermal power output of the heat storage tank; The mathematical model of each energy coupling device in the distributed hydrogen-heat storage system is: (17) Where, H EHP 、 H SOFC are the thermal powers output by the electric heat pump and solid oxide fuel cell respectively; H TST,in is the thermal power absorbed by the heat storage tank; M AE For alkaline electrolysis cells The mass of hydrogen produced in a given time; M HST For hydrogen storage tanks The quality of hydrogen output within a certain time period; L is the lower calorific value of hydrogen; H SOFC is the thermal power output of the solid oxide fuel cell; 、 are the output heat-to-electricity ratio and heat recovery coefficient of the solid oxide fuel cell respectively; Energy conversion efficiency of the equipment; is the heat transfer coefficient.

6. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 1 is characterized in that: The upper-level planning model takes minimizing the annual comprehensive cost of the system as the objective function and is specifically expressed as follows: (18) Where, C eco is the annual economic cost of the system; C re The cost of power supply reliability; C apv Penalty cost for abandoned light; The lower-level planning model takes the minimum daily operating cost of each typical scenario of the distributed hydrogen-heat storage system multi-energy coupling model as the objective function and is specifically expressed as follows: (31) (32) Where, For the scene s Down t The power purchased by the upper power grid at any given moment; for t The electricity purchase price at any given moment; is the time interval; (33) Where, 、 、 They are time-of-use electricity price, heating price and cooling price; For the scene s Down t Solid oxide fuel cell output at all times; For the scene s Down t The heat storage tank supplies heat at all times; For the scene s Down t Cooling load at all times.

7. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 6 is characterized in that: In the upper-level planning model, there are: 1) The annual economic cost of the system is expressed as: (19) Where, C inv is the annualized investment cost of the system; C main is the total annual maintenance cost of the equipment; C ope is the annual operating cost of the system; (20) Where, C inv is the annualized investment cost during the planning period; is the investment cost per unit capacity of the equipment; The equipment investment amount is set to an integer; Invest in rated power for a single device; is the investment cost per unit length of the line; Invest 0-1 variables for the line; For the j The length of the line; The annual investment cost calculation formula considering the equipment capital recovery factor CRF is: (21) Where, r is the interest rate; L The service life of the equipment; The annual equipment maintenance cost of the system considers the annual maintenance cost of each device and transformer in the multi-energy coupling model of the distributed hydrogen-heat storage system, namely: (22) Where, The annual maintenance cost per unit capacity of the equipment; The annual operating cost of the system is the sum of the daily operating costs of the distributed hydrogen-heat storage system multi-energy coupling model in multiple scenarios within a year; (23) Where, The number of days for each typical scenario; For the scene s The daily electricity purchase cost of the multi-energy coupling model of the distributed hydrogen-heat storage system; For the scene s Daily energy sales revenue of the multi-energy coupling model of the distributed hydrogen-heat storage system; 2) Power supply reliability cost in low-probability high-load scenarios The system power restriction index can comprehensively reflect the power outage time and load shortage of each node in the distribution network. The system power restriction index can be converted into power restriction cost as the annual power supply reliability cost of the distribution network to establish the power supply reliability target. C re Expressed as: (24) (25) Where, c p Penalty cost for unit power shortage; E lack The annual power deficit of the system; For the i node t The power shortage at the moment; For the i The existing transformer capacity of the node; For the i Rated capacity of the transformer to be expanded at the node; When the line flow is greater than the line capacity, the line power supply reliability is described by considering the overload power of the distribution network line flow in a low-probability high-load scenario, that is: (26) (27) Where, For the l Overload power of the line; For the l Lines in a low probability high load scenario t Moment power; For the l Existing line capacity of lines; For the l The rated capacity of the line to be expanded; 3) Penalty cost for curtailment in high PV output scenarios Introducing a penalty factor for abandoned solar power and establishing an objective function for abandoned solar power cost in high PV output scenarios C apv , expressed as: (28) (29) (30) Where, c a is the penalty cost per unit of curtailed solar power; The abandoned solar power in high PV output scenarios is caused by insufficient regulation capability of the multi-energy coupling model of the distributed hydrogen-heat storage system; This refers to the abandoned optical power caused by insufficient line capacity in high PV output scenarios; When the system is powered only by photovoltaic power l Active power flowing from the PV node to the load node.

8. The method for planning distribution network expansion with hydrogen-heat storage in a high-proportion photovoltaic scenario according to claim 6 is characterized in that: The constraints are: 1) Equipment capacity constraints of the multi-energy coupling model of distributed hydrogen-heat storage system The rated capacity of the multi-energy coupling model equipment of the planned distributed hydrogen-heat storage system should meet the following constraints: (34) Where, Set a capacity cap for each piece of equipment investment; 2) Energy balance constraints The multi-energy coupling model of the distributed hydrogen-heat storage system should satisfy the multi-energy supply and demand balance of each node, so the planning model should consider the energy balance constraint shown in formula (17); 3) Hydrogen storage tank energy constraints The hydrogen storage tank should meet the hydrogen mass balance requirements: (35) Where, 、 、 、 They are t The quality of residual gas, intake gas, discharge gas and gas loss in hydrogen storage tanks at all times; exist t The hydrogen storage mass at any moment should be less than or equal to the rated capacity of the hydrogen storage tank; (36) Where, is the rated capacity of the hydrogen storage tank; exist t The filling and discharging rate of the hydrogen storage tank should be less than or equal to the rated filling and discharging rate of the hydrogen storage tank at all times; (37) (38) Where, is the rated capacity of the hydrogen storage tank; 4) Thermal storage tank energy constraints The heat storage tank should satisfy the heat balance shown in formula (40) and t The heat storage capacity at any moment should be less than or equal to the rated capacity of the heat storage tank; (39) (40) Where, Q t 、 、 、 They are t Residual heat, input heat, output heat and heat loss of the thermal storage tank at all times; is the rated capacity of the heat storage tank; 5) Solid oxide fuel cell output power constraints Fuel cells in t The electrical power output at any time should be less than or equal to the rated capacity of the fuel cell, expressed as: (41) Where, P SOFC,e is the rated capacity of the solid oxide fuel cell; 6) Transmission line capacity constraints (42) Where, P ij,max 、 Q ij,max for ij Maximum active and reactive power of lines between nodes; S e,ij for ij Rated capacity of inter-node lines; 7) Node voltage constraints The voltage of each node in the distribution network should meet the following constraints: (43) Where, U i For the distribution network i Node voltage; U i,min 、 U i,max Respectively i Node voltage lower and upper limits.

Images