Hydrogen energy storage intelligent soft switch energy scheduling method based on variable efficiency modeling
The intelligent soft-switching energy dispatching method for hydrogen energy storage based on variable efficiency modeling solves the problem of insufficient renewable energy absorption in the distribution network, realizes low-carbon operation and efficient energy management, reduces carbon emissions and improves energy utilization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TIANJIN UNIV
- Filing Date
- 2026-01-08
- Publication Date
- 2026-05-15
AI Technical Summary
Existing technologies have failed to effectively address the multi-timescale consumption problem of renewable energy in the optimized scheduling of energy storage in power distribution networks, and have neglected the variable efficiency characteristics of hydrogen devices, resulting in severe wind and solar curtailment and large carbon emissions.
A smart soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling is adopted. By establishing a model that takes into account the variable efficiency characteristics of hydrogen devices, and combining a low-carbon objective function and operating constraints, a two-layer model predictive control method of efficiency rolling mapping and SOC cut set decomposition is used to carry out energy dispatching at multiple time scales.
It effectively reduces carbon emissions from power distribution network operation, increases the renewable energy absorption rate, improves strategy solution efficiency, and achieves time-coordinated energy management of the system.
Smart Images
Figure CN122052072A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power distribution network energy storage optimization scheduling technology, and in particular, it is a hydrogen energy storage intelligent soft-switching energy scheduling method based on variable efficiency modeling. Background Technology
[0002] In recent years, the high proportion of distributed generation has transformed the traditional operation mode of distribution networks. Distribution networks have become a crucial platform for carrying distributed renewable energy, providing vital support for achieving a green and low-carbon energy transition. However, renewable energy output exhibits significant fluctuations and intermittency across multiple time scales, from intraday to seasonal, severely restricting the distribution network's capacity to absorb renewable energy. Insufficient renewable energy absorption capacity at different time scales leads to severe wind and solar curtailment, low energy utilization, and high carbon dioxide emissions, necessitating multi-time-scale energy dispatching of the distribution network.
[0003] Energy storage systems in distribution networks play a crucial role in achieving energy balance across multiple time scales. Simultaneously, with the rapid development of power electronics technology, intelligent flexible distribution equipment, represented by smart soft switches, has been widely applied in distribution networks. This allows for flexible adjustment of power between feeders, improving system power flow distribution. Smart soft switches have an AC-DC-AC structure, providing a structural foundation for deep coupling of energy storage devices. Energy storage devices can be directly connected to smart soft switches via DC / DC converters, forming an energy storage smart soft switch. This energy storage smart soft switch can control the charging and discharging of energy storage devices while simultaneously controlling power flow, forming a highly integrated energy conversion mode. Energy storage performs energy transfer in the time dimension, while the smart soft switch regulates power in the spatial dimension, achieving spatiotemporal scheduling control in complex source-load scenarios. Compared to the widely used battery energy storage, hydrogen energy storage systems have larger capacity, lower self-discharge losses, and better seasonal energy balance properties. Complementing the regulatory properties of smart soft switches, they form a hydrogen energy storage smart soft switch, better addressing the timing coordination issues of energy management during long-term operation of distribution networks. Furthermore, the energy efficiency of electrolytic hydrogen production devices and hydrogen fuel cells is related to their operating power, and their variable efficiency characteristics need to be considered during energy dispatch, which significantly increases the difficulty of solving the problem.
[0004] Currently, the solutions to the problem of optimal energy storage scheduling in distribution networks primarily rely on smart soft-switching energy dispatch technologies, which often consider battery storage and lack adaptability to seasonal imbalances in the power supply and load of the distribution network. Alternatively, they may consider long-term energy management needs while neglecting the integration of scheduling requirements across multiple time scales, failing to meet the online strategy generation requirements of actual system operation. Furthermore, they may fail to consider the variable efficiency characteristics of electrolytic hydrogen production devices and hydrogen fuel cells, treating the energy efficiency of hydrogen devices as constant, resulting in significant deviations from actual operating conditions. Therefore, there is an urgent need for an energy dispatch method that can meet the online energy dispatch requirements of the system, comprehensively address long-term energy imbalance characteristics, and consider the variable efficiency characteristics of hydrogen devices. This method should be based on variable efficiency modeling and model predictive control to reduce carbon emissions from distribution network operation and improve the renewable energy absorption rate. Summary of the Invention
[0005] The purpose of this invention is to overcome the shortcomings of the prior art and propose a hydrogen energy storage intelligent soft-switching energy dispatch method based on variable efficiency modeling, which can effectively reduce the carbon emissions of the distribution network, improve the renewable energy consumption rate, and improve the strategy solution efficiency.
[0006] The technical problem solved by this invention is achieved through the following technical solution: A smart soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling includes the following steps: Step 1: Based on the selected active distribution network, read the photovoltaic and load forecast data for the scheduling cycle, input the system parameter information, and set the initial time and scheduling cycle for operation scheduling; Step 2: Based on the parameter information in Step 1, establish a smart soft-switching model for hydrogen energy storage that takes into account the variable efficiency characteristics of hydrogen devices, including a smart soft-switching power capacity model, a high-pressure gaseous hydrogen storage tank energy model, a proton exchange membrane hydrogen electrolyzer variable efficiency model, and a proton exchange membrane hydrogen fuel cell variable efficiency model. Step 3: Combine the hydrogen energy storage smart soft switch model that takes into account the variable efficiency characteristics of hydrogen devices obtained in Step 2 with the distribution network operation constraints, and combine it with the low-carbon objective function to establish a long-term low-carbon operation model of the distribution network containing hydrogen energy storage smart soft switch. Step 4: Based on the long-term low-carbon operation model of the distribution network with hydrogen energy storage intelligent soft switch constructed in Step 3, the efficiency rolling mapping method is used to linearize the variable efficiency model of the hydrogen device, and the corresponding solver is called to use the two-layer model predictive control method based on SOC cut set decomposition to solve the scheduling. Step 5: Output the solution results from Step 4, including the power control strategies for proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells during the scheduling period, the energy storage scheduling strategy for high-pressure gaseous hydrogen storage tanks during the scheduling period, and the carbon emissions and photovoltaic absorption rate of the power distribution network during the scheduling period.
[0007] Furthermore, the parameter information input into the system in step 1 includes: annual natural day-level predicted source and load data, intraday hour-level predicted source and load data, distribution network voltage level, topology, and line parameters, connection location, capacity, loss coefficient, and upper limit of power output of the hydrogen energy storage intelligent soft switch, capacity and upper and lower limits of energy storage of the high-pressure gaseous hydrogen storage tank, upper and lower limits of power of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell, ramping parameters, and average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell. The value of the carbon emission factor for electricity.
[0008] Furthermore, the intelligent soft-switching power capacity model in step 2 is as follows: in, , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , The active power; , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , reactive power; , Intelligent soft switching for hydrogen energy storage across time cross-sections At the distribution network node , Power loss generated by the converter; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The loss coefficient of the converter; Time section The active power output of a proton exchange membrane hydrogen fuel cell; Time section The active power input to the proton exchange membrane hydrogen electrolyzer; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The capacity of the converter.
[0009] Furthermore, the energy model for the high-pressure gaseous hydrogen storage tank in step 2 is as follows: in, , The upper and lower limits of the active power output of a proton exchange membrane hydrogen fuel cell; , The upper and lower limits of active power input for proton exchange membrane hydrogen electrolyzers; Time section Proton exchange membrane hydrogen fuel cell output active power ramp-up; , The upper and lower limits of the ramp-up of active power output for proton exchange membrane hydrogen fuel cells; Time section Input active power ramp-up of proton exchange membrane hydrogen electrolyzer; , Upper and lower limits of active power ramp-up for proton exchange membrane hydrogen electrolyzers; Time section The amount of hydrogen stored in the hydrogen storage tank; , The upper and lower limits of hydrogen mass stored in hydrogen storage tanks; For time cross section ( The hydrogen storage tank stores the mass of hydrogen gas. Time section Mass of hydrogen produced by a proton exchange membrane hydrogen electrolyzer; Time section The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell; For a single time segment interval; Time section Net hydrogen intake; This is the integration flag; It is Faraday's constant; For Faraday efficiency; This refers to the molar mass of hydrogen gas. Time section Current in a proton exchange membrane hydrogen electrolyzer; Time section Proton exchange membrane hydrogen fuel cell current; Number of proton exchange membrane hydrogen production electrolyzers; This refers to the number of proton exchange membrane hydrogen fuel cells; Time cross section Energy efficiency of proton exchange membrane hydrogen electrolyzer; Time section Energy efficiency of proton exchange membrane hydrogen fuel cells; Hydrogen has a low calorific value; It has a high calorific value, which is due to hydrogen.
[0010] Furthermore, the variable efficiency model for the proton exchange membrane hydrogen production electrolyzer in step 2 is as follows: in, , Maximum voltage and current values for a proton exchange membrane hydrogen electrolyzer; Equivalent resistance for a proton exchange membrane hydrogen electrolyzer; , is the fitting constant.
[0011] Furthermore, the variable efficiency model for the proton exchange membrane hydrogen production electrolyzer in step 2 is as follows: in, Time section Proton exchange membrane hydrogen fuel cell voltage; Time section Open-circuit voltage of a proton exchange membrane hydrogen fuel cell; Time section Operating pressure drop of proton exchange membrane hydrogen fuel cells; Time section Ohmic pressure drop in a proton exchange membrane hydrogen fuel cell; Time section Polymerization pressure drop in proton exchange membrane hydrogen fuel cells; For Gibbs free energy; It is an entropy change; , For the operating temperature and reference temperature of proton exchange membrane hydrogen fuel cells; It is the gas constant; Time section Hydrogen partial pressure; Time section Oxygen partial pressure; Time section The amount of oxygen polymerized on the surface of the cathode catalyst; The equivalent resistance of a proton exchange membrane hydrogen fuel cell; This is the concentration overpotential coefficient; This represents the maximum current of a proton exchange membrane hydrogen fuel cell. , , , is the fitting constant.
[0012] Furthermore, the distribution network operation constraints in step 3 are: in, A collection of distribution network lines; branch road The resistance; branch road The reactance; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node The current; Time section node Flow to Node The current; Time section node The sum of the active power injected upwards; Time section node The sum of reactive power injected upwards; Time section node The voltage amplitude; Time section node The voltage amplitude; and They are time sections node The reactive power consumed by photovoltaic power and loads; and They are time sections node The active power consumed by photovoltaic power and loads; Time section The active power transmitted from the main grid to the distribution network; and These are the minimum and maximum allowable node voltage amplitudes for the distribution network, respectively. This represents the maximum allowable branch current amplitude in the distribution network.
[0013] Furthermore, the low-carbon objective function in step 3 is: in, For long-term carbon emissions from power distribution networks; Carbon emissions resulting from electricity consumption in the power distribution network; Carbon emissions resulting from energy loss due to intelligent soft switching in hydrogen energy storage; This refers to the reduction in carbon emissions resulting from photovoltaic power generation. To calculate the total number of time sections; This represents the total number of time segments within the scheduling cycle. Carbon emission factor for electricity; For distribution network nodes; Time section node Active load at the location; A collection of distribution network lines; For the line The resistance; Time section line Current on; A set of nodes that connect to intelligent soft switches in the distribution network; A collection of energy storage devices; Time section Energy storage devices Energy loss generated; A collection of photovoltaic access nodes; Time cross section node The amount of active power consumed by photovoltaic power generation.
[0014] Furthermore, the efficiency rolling mapping method in step 4 includes the following steps: Step 4.1.1: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switch containing hydrogen energy storage, combined with the annual natural day-level predicted source-load data, the average energy efficiency sequence of proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in Step 1, The daily offline scheduling plan for hydrogen storage was calculated recently. ; Step 4.1.2: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switch containing hydrogen energy storage, combined with the intraday hourly predicted source-load data, the average energy efficiency sequence of proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in Step 1, The daily hydrogen storage offline scheduling plan generated in step 4.1.1 Under constraints, the offline power control strategy for the hydrogen unit at the hourly level within the day was calculated; Step 4.1.3: Based on the function mapping relationship obtained from the intelligent soft-switching model of hydrogen energy storage considering the variable efficiency characteristics of hydrogen devices. The intraday hourly offline power control strategy for hydrogen devices generated in step 4.1.2 is mapped to the average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in step 1. In this process, the efficiency pre-mapped sequence is obtained. ; Step 4.1.4: Update the efficiency pre-mapping sequence in real time based on the actual power and efficiency conditions of the hydrogen unit during operation. This reflects the actual operating conditions, resulting in an efficiency rolling mapping matrix, which is then applied to the real-time scheduling and control process. The efficiency rolling mapping matrix is expressed as follows: in, For the first Intra-day efficiency rolling mapping matrix; For the first The efficiency pre-mapping sequence for the first hour of the day; For the first The efficiency pre-mapped sequence for the second hour of the day; For the first The efficiency pre-mapped sequence for the 24th hour of the day; For the first The efficiency rolling mapping sequence for the first hour of the day; For the first The efficiency rolling mapping sequence for the second hour of the day.
[0015] Furthermore, the scheduling solution in step 4, based on the two-level model predictive control method using SOC cut set decomposition, includes the following steps: Step 4.2.1: Set the scheduling period. The upper-level model's predictive control time granularity is the calendar day, and the lower-level model's predictive control time granularity is the hour. Set the scheduling control time to [time value missing]. ,in, To schedule the number of days at a given time point in real time, To set the hour for real-time scheduling, the initial time for operation and scheduling is set. , ; Step 4.2.2: Based on the input annual daily-level predicted source-load data and the established long-term low-carbon operation model of the distribution network with hydrogen energy storage intelligent soft switching, call the solver to generate an annual hydrogen energy storage scheduling plan with daily-level time granularity, as the first... TianSOC cut set ,in, For the first Net hydrogen inventory per day For the first The mass of hydrogen produced by the proton exchange membrane hydrogen electrolyzer; For the first The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell. For the first Heavenly Net hydrogen intake per hour The interval is a natural day time segment. The time interval is the hourly cross-section. This represents the number of hours within a day, with a value of 24, and is continuously updated and calculated based on the scheduling process at the natural day time granularity. Step 4.2.3: Based on step 4.2.2, the generated first... TianSOC cut set Combining the input intraday hourly predicted source-load data, the established long-term low-carbon operation model of the distribution network with hydrogen-containing energy storage intelligent soft switch, and the efficiency rolling mapping matrix The solver is invoked to generate the daily power output of a proton exchange membrane hydrogen electrolyzer and a proton exchange membrane hydrogen fuel cell with hourly time granularity. , Control strategies are implemented and updated on a rolling basis at the hourly time granularity within the day. Step 4.2.4: Determine whether the daily scheduling plan for all time periods has been generated. If not, update... for Then return to step 4.2.3; if so, update the daily distribution network carbon emissions and photovoltaic absorption rate data, update the high-pressure gaseous hydrogen storage tank energy storage data, and update... Proceed to the next step; Step 4.2.5: Determine whether the generation of scheduling control plans for all time periods of the scheduling cycle has been completed. If not, update... for Then return to step 4.2.2; if so, end.
[0016] The advantages and positive effects of this invention are: This invention addresses the problem of optimizing energy storage scheduling in distribution networks. It fully considers the system's online energy scheduling requirements, the long-term energy imbalance characteristics of the distribution network, and the variable efficiency characteristics of hydrogen storage devices. It establishes a smart soft-switching model for hydrogen storage that takes into account the variable efficiency characteristics of hydrogen storage devices, and a long-term low-carbon operation model for the distribution network incorporating this model. The variable efficiency model of the hydrogen storage device is linearized using an efficiency rolling mapping method. A two-layer model predictive control method based on SOC cut-set decomposition is employed, and relevant mathematical solvers are called to solve the problem, yielding a long-term online scheduling strategy for hydrogen storage. Using the hydrogen storage smart soft-switching energy scheduling method based on variable efficiency modeling, as described in this invention, to optimize the scheduling of energy storage in distribution networks can effectively reduce carbon emissions from distribution network operation, improve the renewable energy absorption rate, and simultaneously improve the efficiency of strategy solution. Attached Figure Description
[0017] Figure 1 This is a flowchart of the intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling, as described in this invention. Figure 2 This invention provides electrical load and photovoltaic forecast / actual data for various time scales. Figure 3 This is a schematic diagram of the intelligent soft switch structure for hydrogen energy storage of the present invention; Figure 4 This is a power distribution network topology diagram for the example of this invention; Figure 5 This invention provides an online power control strategy for a year-round proton exchange membrane hydrogen electrolyzer and a proton exchange membrane hydrogen fuel cell. Figure 6 This invention provides a comparison of the annual energy storage scheduling strategies for high-pressure gaseous hydrogen storage tanks. Figure 7 The annual photovoltaic power consumption data for different schemes of this invention; Figure 8 This is a typical intraday efficiency rolling mapping result for a hydrogen unit according to the present invention; Figure 9 This invention provides a typical intraday online power control strategy for proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells. Figure 10 This paper compares the power control strategies of hydrogen devices obtained by different methods within the typical daily set of this invention. Detailed Implementation
[0018] The present invention will be further described in detail below with reference to the accompanying drawings.
[0019] A smart soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling, such as Figure 1 As shown, it includes the following steps: Step 1: Based on the selected active distribution network, read the photovoltaic and load forecast data for the scheduling cycle, input the system parameter information, and set the initial time and scheduling cycle for operation scheduling.
[0020] The input system parameters include: annual daily predicted source-load data, intraday hourly predicted source-load data, distribution network voltage level, topology, and line parameters; connection location, capacity, loss coefficient, and upper power output limit of the hydrogen storage smart soft switch; capacity and upper and lower limits of the high-pressure gaseous hydrogen storage tank; upper and lower limits of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell power, ramping parameters, and average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell. The value of the carbon emission factor for electricity.
[0021] Step 2: Based on the parameter information provided in Step 1, establish a smart soft-switching model for hydrogen energy storage that considers the variable efficiency characteristics of the hydrogen device. This includes a smart soft-switching power-capacity model, a high-pressure gaseous hydrogen storage tank energy model, a proton exchange membrane hydrogen electrolyzer variable efficiency model, and a proton exchange membrane hydrogen fuel cell variable efficiency model. These models respectively consider the power, capacity, and loss characteristics of the smart soft-switching model, the operating characteristics of the high-pressure gaseous hydrogen storage tank, and the operating and variable efficiency characteristics of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell. The intelligent soft-switching power capacity model is as follows: (1) (2) (3) (4) (5) in, , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , The active power; , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , reactive power; , Intelligent soft switching for hydrogen energy storage across time cross-sections At the distribution network node , Power loss generated by the converter; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The loss coefficient of the converter; Time section The active power output of a proton exchange membrane hydrogen fuel cell; Time section The active power input to the proton exchange membrane hydrogen electrolyzer; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The capacity of the converter.
[0022] The energy model for a high-pressure gaseous hydrogen storage tank is as follows: (6) (7) (8) (9) in, , The upper and lower limits of the active power output of a proton exchange membrane hydrogen fuel cell; , The upper and lower limits of active power input for proton exchange membrane hydrogen electrolyzers; Time section Proton exchange membrane hydrogen fuel cell output active power ramp-up; , The upper and lower limits of the ramp-up of active power output for proton exchange membrane hydrogen fuel cells; Time section Input active power ramp-up of proton exchange membrane hydrogen electrolyzer; , Upper and lower limits of active power ramp-up for proton exchange membrane hydrogen electrolyzers; Time section The amount of hydrogen stored in the hydrogen storage tank; , The upper and lower limits of hydrogen mass stored in hydrogen storage tanks; For time cross section ( The hydrogen storage tank stores the mass of hydrogen gas. Time section Mass of hydrogen produced by a proton exchange membrane hydrogen electrolyzer; Time section The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell; For a single time segment interval; Time section Net hydrogen intake; This is the integration flag; It is Faraday's constant; For Faraday efficiency; This refers to the molar mass of hydrogen gas. Time section Current in a proton exchange membrane hydrogen electrolyzer; Time section Proton exchange membrane hydrogen fuel cell current; Number of proton exchange membrane hydrogen production electrolyzers; This refers to the number of proton exchange membrane hydrogen fuel cells; Time cross section Energy efficiency of proton exchange membrane hydrogen electrolyzer; Time section Energy efficiency of proton exchange membrane hydrogen fuel cells; Hydrogen has a low calorific value; It has a high calorific value, which is due to hydrogen.
[0023] The variable efficiency model for a proton exchange membrane hydrogen electrolyzer is as follows: (10) (11) (12) in, , Maximum voltage and current values for a proton exchange membrane hydrogen electrolyzer; Equivalent resistance for a proton exchange membrane hydrogen electrolyzer; , is the fitting constant.
[0024] The variable efficiency model for proton exchange membrane hydrogen fuel cells is as follows: (13) (14) (15) (16) (17) (18) (19) in, Time section Proton exchange membrane hydrogen fuel cell voltage; Time section Open-circuit voltage of a proton exchange membrane hydrogen fuel cell; Time section Operating pressure drop of proton exchange membrane hydrogen fuel cells; Time section Ohmic pressure drop in a proton exchange membrane hydrogen fuel cell; Time section Polymerization pressure drop in proton exchange membrane hydrogen fuel cells; For Gibbs free energy; It is an entropy change; , For the operating temperature and reference temperature of proton exchange membrane hydrogen fuel cells; It is the gas constant; Time section Hydrogen partial pressure; Time section Oxygen partial pressure; Time section The amount of oxygen polymerized on the surface of the cathode catalyst; The equivalent resistance of a proton exchange membrane hydrogen fuel cell; This is the concentration overpotential coefficient; This represents the maximum current of a proton exchange membrane hydrogen fuel cell. , , , is the fitting constant.
[0025] Step 3: Combine the hydrogen energy storage intelligent soft-switching model considering the variable efficiency characteristics of the hydrogen device obtained in Step 2 with the distribution network operation constraints, and combine it with the low-carbon objective function to establish a long-term low-carbon operation model for the distribution network containing hydrogen energy storage intelligent soft-switching; among which, Step 3.1, Distribution network operation constraints, are expressed as: (20) (twenty one) (twenty two) (twenty three) (twenty four) (25) (26) (27) (28) in, A collection of distribution network lines; branch road The resistance; branch road The reactance; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node The current; Time section node Flow to Node The current; Time section node The sum of the active power injected upwards; Time section node The sum of reactive power injected upwards; Time section node The voltage amplitude; Time section node The voltage amplitude; and They are time sections node The reactive power consumed by photovoltaic power and loads; and They are time sections node The active power consumed by photovoltaic power and loads; Time section The active power transmitted from the main grid to the distribution network; and These are the minimum and maximum allowable node voltage amplitudes for the distribution network, respectively. This represents the maximum allowable branch current amplitude in the distribution network.
[0026] Step 3.2: Construct the low-carbon objective function, expressed as: (29) (30) (31) (32) in, For long-term carbon emissions from power distribution networks; Carbon emissions resulting from electricity consumption in the power distribution network; Carbon emissions resulting from energy loss due to intelligent soft switching in hydrogen energy storage; This refers to the reduction in carbon emissions resulting from photovoltaic power generation. To calculate the total number of time sections; This represents the total number of time segments within the scheduling cycle. Carbon emission factor for electricity; For distribution network nodes; A set of nodes that connect to intelligent soft switches in the distribution network; A collection of energy storage devices; Time section Energy storage devices Energy loss generated; This is a collection of photovoltaic access nodes.
[0027] Step 3.3, the long-term low-carbon operation model of the distribution network with hydrogen-containing energy storage intelligent soft switch, can be written in compact form as follows: (33) Step 4: Based on the long-term low-carbon operation model of the distribution network with hydrogen energy storage intelligent soft switch constructed in Step 3, the efficiency rolling mapping method is used to linearize the variable efficiency model of the hydrogen device, and the corresponding solver is called to use the two-layer model predictive control method based on SOC cut set decomposition to solve the scheduling. The efficiency-driven rolling mapping method includes the following steps: Step 4.1.1, Pre-generation of Daily Hydrogen Storage Dispatch Plan: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switch containing hydrogen storage, combined with the annual natural day-level predicted source-load data, and the average energy efficiency sequence of proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in Step 1, The daily offline scheduling plan for hydrogen storage was calculated recently. ; Step 4.1.2: Pre-generation of intraday hourly power control strategy for hydrogen devices: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switching for hydrogen energy storage, combined with the intraday hourly predicted source-load data, average energy efficiency sequences of proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells input in Step 1, The daily hydrogen storage offline scheduling plan generated in step 4.1.1 Under constraints, the offline power control strategy for the hydrogen unit at the hourly level within the day was calculated; The intraday hourly offline power control strategy for hydrogen units is expressed as follows: (34) in, For the first Daily hourly-level offline power control strategy for hydrogen plants; For the first Offline power control strategy for the first hour of the hydrogen unit; For the first The 24-hour offline power control strategy for the hydrogen unit.
[0028] Step 4.1.3, Efficiency Sequence Pre-mapping: Based on the function mapping relationship obtained from the intelligent soft-switching model of hydrogen energy storage considering the variable efficiency characteristics of hydrogen devices. The intraday hourly offline power control strategy for hydrogen devices generated in step 4.1.2 is mapped to the average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and the proton exchange membrane hydrogen fuel cell input in step 1. In this process, the efficiency pre-mapped sequence is obtained. ; Function mapping relationship Represented as: (35) in, This is a mapping direction flag; it is a sequence update conversion flag.
[0029] Efficiency premapped sequence Represented as: (36) in, For the first Efficient pre-mapped sequence; For the first The efficiency pre-mapping sequence for the first hour of the day; For the first The efficiency pre-mapped sequence for the second hour of the day; For the first The efficiency pre-mapped sequence for the 24th hour of the day.
[0030] Step 4.1.4, Online calibration of efficiency mapping sequence: Update the efficiency pre-mapping sequence in real time based on the actual power and efficiency conditions of the hydrogen unit during operation. This reflects the actual operating conditions, yields an efficiency rolling mapping matrix, and is applied to the real-time scheduling and control process. The efficiency rolling mapping matrix is represented as: (37) in, For the first Intra-day efficiency rolling mapping matrix; For the first The efficiency rolling mapping sequence for the first hour of the day; For the first The efficiency rolling mapping sequence for the second hour of the day.
[0031] The two-level model predictive control method based on SOC cut set decomposition includes the following steps: Step 4.2.1: Set the scheduling cycle parameters and the time granularity of the upper and lower level model predictive control: Set the scheduling cycle to 1 year, the time granularity of the upper level model predictive control to calendar days, and the time granularity of the lower level model predictive control to hours. Set the scheduling control time to... ,in, To schedule the number of days at a given time point in real time, To set the hour for real-time scheduling, the initial time for operation and scheduling is set. , ; Step 4.2.2, Upper-level model predictive control rolling calculation: Based on the input annual natural day-level predicted source-load data and the established long-term low-carbon operation model of the distribution network containing hydrogen energy storage intelligent soft switches, the solver is invoked to generate an annual hydrogen energy storage scheduling plan with natural day-level time granularity, as the first... TianSOC cut set ,in, For the first Net hydrogen inventory per day For the first The mass of hydrogen produced by the proton exchange membrane hydrogen electrolyzer; For the first The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell. For the first Heavenly Net hydrogen intake per hour The interval is a natural day time segment. The time interval is the hourly cross-section. This represents the number of hours within a day, with a value of 24, and is continuously updated and calculated based on the scheduling process at the natural day time granularity. The rolling update of data at the annual calendar day time granularity is represented as follows: (38) in, The dataset is updated on a rolling basis throughout the year; The dataset is updated on a rolling basis for day 1; The dataset is updated on a rolling basis for the second day; The dataset is updated on a rolling basis every 365 days. This is the prediction dataset for day 1; This is the predicted dataset for day 2; This is the predicted dataset for day 365. This is the data set from day 1. This is the data set from the measurements taken on day 2.
[0032] Step 4.2.3, Lower-level model predictive control rolling calculation: Based on step 4.2.2, the generated... TianSOC cut set Combining the input intraday hourly predicted source-load data, the established long-term low-carbon operation model of the distribution network with hydrogen-containing energy storage intelligent soft switch, and the efficiency rolling mapping matrix The solver is invoked to generate the daily power output of a proton exchange membrane hydrogen electrolyzer and a proton exchange membrane hydrogen fuel cell with hourly time granularity. , Control strategies are implemented and updated on a rolling basis at the hourly time granularity within the day. The rolling update of data at the hourly time granularity within a day is represented as follows: (39) in, For the first The dataset is updated daily. For the first The dataset is updated on a rolling basis every hour of the day. For the first The dataset is updated every second hour of the day. For the first The dataset is updated every 24 hours. For the first Prediction dataset for the first hour of the day; For the first Prediction dataset for the second hour of the day; For the first Predicted dataset for the 24th hour of the day; For the first Data set of measurements taken in the first hour of the day; For the first The second hour of measurement data.
[0033] Step 4.2.4: Determine whether the daily scheduling plan for all time periods has been generated. If not, update... for (40) Then return to step 4.2.3; if so, update the daily distribution network carbon emissions and photovoltaic absorption rate data, update the high-pressure gaseous hydrogen storage tank energy storage data, and update... Proceed to the next step; Step 4.2.5: Determine whether the generation of scheduling control plans for all time periods of the scheduling cycle has been completed. If not, update... for (41) Then return to step 4.2.2; if so, end.
[0034] Step 5: Output the solution results from Step 4, including the power control strategies for proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells during the scheduling period, the energy storage scheduling strategy for high-pressure gaseous hydrogen storage tanks during the scheduling period, and the carbon emissions and photovoltaic absorption rate of the power distribution network during the scheduling period.
[0035] Based on the above-mentioned intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling, the effectiveness of the invention was verified through actual testing.
[0036] First, enter as follows: Figure 2The system displays the predicted and actual load curves at different time scales within the scheduling period, as well as the predicted and actual photovoltaic output curves. Then, it inputs system parameters such as the distribution network voltage level, topology, line parameters, the connection location, capacity, loss coefficient, and upper power output limit of the hydrogen storage intelligent soft switch, the capacity and upper and lower limits of the high-pressure gaseous hydrogen storage tank, the upper and lower limits of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell power, ramp parameters, average energy efficiency, and the value of the electricity carbon emission factor. Finally, it sets the initial scheduling time and scheduling period. The structure of the hydrogen storage intelligent soft switch is shown below. Figure 3 As shown. Figure 4 In the power distribution network shown, the voltage level is 12.66kV, the total active load is 4885kW, and the total reactive load is 2710kvar. Two intelligent soft switches for hydrogen energy storage are connected to nodes 55 and 95 and nodes 117 and 123, respectively. The converter capacity at each port is 2MVA, the loss factor is 0.01, and the per-unit voltage allowable range is [0.95, 1.05]. The high-pressure gaseous hydrogen storage capacity is 100MWh, the power range of the proton exchange membrane hydrogen electrolyzer is [0.05, 1]MW, the power range of the proton exchange membrane hydrogen fuel cell is [0.05, 1]MW, the equipment ramp-up limit is 20% of the maximum equipment power, and the average energy efficiency of the electrolyzer and fuel cell is taken as 80% and 55%, respectively. Detailed parameters are shown in Table 1. The annual daily predicted source load data is subject to a 30% random error, and the intraday hourly predicted source load data is subject to a 10% random error. The carbon emission factor for electricity is taken as the national average carbon dioxide emission factor for electricity in 2024, which is 0.5777 kgCO2 / kWh. The dispatch cycle is set to 1 year.
[0037] Table 1 System Composition and Parameters
[0038] The computer hardware environment for performing the optimized calculations is an Intel Core i7-13700 with a clock speed of 2.10GHz and 16GB of memory, and the software environment is a Windows 11 operating system.
[0039] To verify the effectiveness of the method proposed in this invention in long-term energy scheduling, for Figure 4 The annual carbon emissions and photovoltaic power consumption indicators of the power distribution network shown are compared and controlled using the following three schemes: Option 1: Initial year-round scenario, no smart soft switches are connected to the distribution network; Option II: Two sets of hydrogen energy storage smart soft switches are connected to the distribution network, connecting nodes 55 and 95 and nodes 117 and 123 respectively. The model predictive control offline calculation is used to obtain the annual offline strategy for the actual operation of the distribution network. Option III: Two sets of intelligent soft switches for hydrogen energy storage are connected to the distribution network, connecting nodes 55 and 95 and nodes 117 and 123 respectively. The efficiency rolling mapping method and the two-layer model predictive control method based on SOC cut set decomposition proposed in this invention are used to obtain the annual online scheduling strategy for the actual operation of the distribution network.
[0040] The calculated annual online power control strategies for proton exchange membrane hydrogen electrolyzers and proton exchange membrane fuel cells, on a daily basis, are shown below. Figure 5 A comparison of energy storage scheduling strategies for high-pressure gaseous hydrogen storage tanks throughout the year can be found in [link to relevant documentation]. Figure 6 See below for the annual photovoltaic power consumption figures for different schemes. Figure 7 Table 2 shows the annual carbon emissions and photovoltaic absorption rate for different schemes.
[0041] Table 2 Annual carbon emissions and photovoltaic grid integration rate for different schemes
[0042] from Figure 5 , Figure 6 The annual hydrogen storage dispatch strategy and Figure 7 Table 2 shows a comparison of annual carbon emissions and photovoltaic (PV) integration rates for different schemes. It can be seen that this invention considers the annual source-load imbalance and reduces the annual carbon emissions of the distribution network and improves the overall annual PV integration rate through the year-round scheduling of hydrogen device energy. Furthermore, the method proposed in this invention uses online calculation with rolling data updates, resulting in lower annual carbon emissions and a higher PV integration rate compared to offline calculations using only predicted data.
[0043] To verify the effectiveness of the method proposed in this invention in intraday energy scheduling, for Figure 4 The typical daily carbon emissions and photovoltaic absorption indicators of the distribution network shown are compared and controlled using the following four schemes: Option IV: The daily scheduling plan is evenly distributed, and no hourly scheduling is carried out for the hydrogen units; Option V: The intraday offline power control strategy calculated based on the average efficiency of the hydrogen unit and the intraday source load forecast data is used for the actual operation throughout the day; Option VI: Based on the average efficiency of the hydrogen unit, the power generation control strategy is continuously optimized in real time for use throughout the day's actual operation. Scheme VII: The real-time power control strategy for the hydrogen unit, obtained by adopting the efficiency rolling mapping method and the two-layer model predictive control method based on SOC cut set decomposition proposed in this invention, is used for the actual operation process throughout the day.
[0044] The calculated rolling mapping results of typical intraday hydrogen unit efficiency in hours are shown below. Figure 8For typical intraday online power control strategies of proton exchange membrane hydrogen production electrolyzers and proton exchange membrane hydrogen fuel cells, see [link to relevant documentation]. Figure 9 Table 3 shows the typical daily carbon emissions and photovoltaic absorption rate for different schemes.
[0045] Table 3 Typical Daily Carbon Emissions and Photovoltaic Grid Integration Rates for Different Schemes
[0046] from Figure 9 The typical intraday online power control strategies for proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells shown in Table 3, along with the comparison of typical intraday carbon emissions and photovoltaic absorption rates for different schemes, demonstrate that this invention considers intraday source-load fluctuations and the variable efficiency characteristics of the hydrogen device. Through intraday energy scheduling of the hydrogen device, it reduces intraday carbon emissions from the distribution network and improves the intraday photovoltaic absorption rate of the distribution network. Furthermore, the calculation results of the method proposed in this invention are significantly better than calculation methods that do not perform intraday scheduling, use intraday offline scheduling, or treat the efficiency of the hydrogen device as constant.
[0047] To verify the effectiveness of the method proposed in this invention in improving computational efficiency, [the following is conducted]: Figure 4 The carbon emissions and photovoltaic absorption capacity of a typical daily distribution network over 288 hours are compared and controlled using the following three schemes: Scheme VIII: Initial typical daily set scenario, no smart soft switch connected to the distribution network; Option IX: Two sets of intelligent soft switches for hydrogen energy storage are connected to the distribution network, connecting nodes 55 and 95 and nodes 117 and 123 respectively. A centralized optimization method is used to obtain a typical daily hydrogen device scheduling strategy for distribution network operation. Option X: Two sets of intelligent soft switches for hydrogen energy storage are connected to the distribution network, connecting nodes 55 and 95 and nodes 117 and 123 respectively. The efficiency rolling mapping method and the two-layer model predictive control method based on SOC cut set decomposition proposed in this invention are used to obtain a typical daily hydrogen unit scheduling strategy for distribution network operation.
[0048] A comparison of the hydrogen plant power control strategies calculated by the method of this invention within a typical daily set of centralized optimizations and those calculated on an hourly basis is shown in the figure. Figure 10 Table 4 shows a comparison of the typical day set solution results for different schemes.
[0049] Table 4 Comparison of typical day set solution results from different schemes
[0050] from Figure 10The comparison between the typical daily set optimization and the hourly hydrogen device power control strategy calculated by the method of this invention, as shown in Table 4, and the comparison of the typical daily set solution results of different schemes, show that the method proposed in this invention, compared with the traditional centralized optimization method, significantly shortens the calculation time and improves the calculation efficiency, meeting the online energy dispatch requirements of the system, under the premise of a carbon emission reduction deviation of 1.70% and a photovoltaic absorption rate deviation of 0.12%.
[0051] In summary, the intelligent soft-switching energy dispatching method for hydrogen energy storage based on variable efficiency modeling proposed in this invention fully considers the variable efficiency characteristics of proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells, as well as the energy imbalance characteristics of the distribution network at different time scales. It effectively realizes intelligent soft-switching energy dispatching of hydrogen energy storage in the distribution network, reduces carbon emissions from the distribution network, improves photovoltaic absorption rate, and enhances strategy solution efficiency to meet the online energy dispatching requirements of the system.
[0052] It should be emphasized that the embodiments described in this invention are illustrative rather than limiting. Therefore, this invention includes, but is not limited to, the embodiments described in the specific implementation. Any other implementations derived by those skilled in the art based on the technical solutions of this invention are also within the scope of protection of this invention.
Claims
1. A smart soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling, characterized in that, Includes the following steps: Step 1: Based on the selected active distribution network, read the photovoltaic and load forecast data for the scheduling cycle, input the system parameter information, and set the initial time and scheduling cycle for operation scheduling; Step 2: Based on the parameter information in Step 1, establish a smart soft-switching model for hydrogen energy storage that takes into account the variable efficiency characteristics of hydrogen devices, including a smart soft-switching power capacity model, a high-pressure gaseous hydrogen storage tank energy model, a proton exchange membrane hydrogen electrolyzer variable efficiency model, and a proton exchange membrane hydrogen fuel cell variable efficiency model. Step 3: Combine the hydrogen energy storage smart soft switch model that takes into account the variable efficiency characteristics of hydrogen devices obtained in Step 2 with the distribution network operation constraints, and combine it with the low-carbon objective function to establish a long-term low-carbon operation model of the distribution network containing hydrogen energy storage smart soft switch. Step 4: Based on the long-term low-carbon operation model of the distribution network with hydrogen energy storage intelligent soft switch constructed in Step 3, the efficiency rolling mapping method is used to linearize the variable efficiency model of the hydrogen device, and the corresponding solver is called to use the two-layer model predictive control method based on SOC cut set decomposition to solve the scheduling. Step 5: Output the solution results from Step 4, including the power control strategies for proton exchange membrane hydrogen electrolyzers and proton exchange membrane hydrogen fuel cells during the scheduling period, the energy storage scheduling strategy for high-pressure gaseous hydrogen storage tanks during the scheduling period, and the carbon emissions and photovoltaic absorption rate of the power distribution network during the scheduling period.
2. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The parameter information input into the system in step 1 includes: annual daily predicted source-load data, intraday hourly predicted source-load data, distribution network voltage level, topology, and line parameters; connection location, capacity, loss coefficient, and upper power output limit of the hydrogen energy storage smart soft switch; capacity and upper and lower limits of the high-pressure gaseous hydrogen storage tank; upper and lower limits of the power of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell, ramping parameters, and average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell. The value of the carbon emission factor for electricity.
3. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The intelligent soft-switching power capacity model in step 2 is as follows: ； ； ； ； ； in, , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , The active power; , Intelligent soft switching for hydrogen energy storage across time cross-sections Injection into distribution network nodes , reactive power; , Intelligent soft switching for hydrogen energy storage across time cross-sections At the distribution network node , Power loss generated by the converter; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The loss coefficient of the converter; Time section The active power output of a proton exchange membrane hydrogen fuel cell; Time section The active power input to the proton exchange membrane hydrogen electrolyzer; , Intelligent soft switches for hydrogen energy storage at distribution network nodes , The capacity of the converter.
4. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The energy model for the high-pressure gaseous hydrogen storage tank in step 2 is as follows: ； ； ； ； in, , The upper and lower limits of the active power output of a proton exchange membrane hydrogen fuel cell; , The upper and lower limits of active power input for proton exchange membrane hydrogen electrolyzers; Time section Proton exchange membrane hydrogen fuel cell output active power ramp-up; , The upper and lower limits of the ramp-up of active power output for proton exchange membrane hydrogen fuel cells; Time section Input active power ramp-up of proton exchange membrane hydrogen electrolyzer; , Upper and lower limits of active power ramp-up for proton exchange membrane hydrogen electrolyzers; Time section The amount of hydrogen stored in the hydrogen storage tank; , The upper and lower limits of hydrogen mass stored in hydrogen storage tanks; For time cross section ( The hydrogen storage tank stores the mass of hydrogen gas. Time section Mass of hydrogen produced by a proton exchange membrane hydrogen electrolyzer; Time section The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell; For a single time segment interval; Time section Net hydrogen intake; This is the integration flag; It is Faraday's constant; For Faraday efficiency; This refers to the molar mass of hydrogen gas. Time section Current in a proton exchange membrane hydrogen electrolyzer; Time section Proton exchange membrane hydrogen fuel cell current; Number of proton exchange membrane hydrogen production electrolyzers; This refers to the number of proton exchange membrane hydrogen fuel cells; Time cross section Energy efficiency of proton exchange membrane hydrogen electrolyzer; Time section Energy efficiency of proton exchange membrane hydrogen fuel cells; Hydrogen has a low calorific value; It has a high calorific value, which is due to hydrogen.
5. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The variable efficiency model for the proton exchange membrane hydrogen production electrolyzer in step 2 is as follows: ； ； ； in, , Maximum voltage and current values for a proton exchange membrane hydrogen electrolyzer; Equivalent resistance for a proton exchange membrane hydrogen electrolyzer; , is the fitting constant.
6. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The variable efficiency model for the proton exchange membrane hydrogen fuel cell in step 2 is as follows: ； ； ； ； ； ； ； in, Time section Proton exchange membrane hydrogen fuel cell voltage; Time section Open-circuit voltage of a proton exchange membrane hydrogen fuel cell; Time section Operating pressure drop of proton exchange membrane hydrogen fuel cells; Time section Ohmic pressure drop in a proton exchange membrane hydrogen fuel cell; Time section Polymerization pressure drop in proton exchange membrane hydrogen fuel cells; For Gibbs free energy; It is an entropy change; , For the operating temperature and reference temperature of proton exchange membrane hydrogen fuel cells; It is the gas constant; Time section Hydrogen partial pressure; Time section Oxygen partial pressure; Time section The amount of oxygen polymerized on the surface of the cathode catalyst; The equivalent resistance of a proton exchange membrane hydrogen fuel cell; This is the concentration overpotential coefficient; This represents the maximum current of a proton exchange membrane hydrogen fuel cell. , , , is the fitting constant.
7. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The power distribution network operation constraints in step 3 are: ； ； ； ； ； ； ； ； ； in, A collection of distribution network lines; branch road The resistance; branch road The reactance; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node The active power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node reactive power; Time section node Flow to Node The current; Time section node Flow to Node The current; Time section node The sum of the active power injected upwards; Time section node The sum of reactive power injected upwards; Time section node The voltage amplitude; Time section node The voltage amplitude; and They are time sections node The reactive power consumed by photovoltaic power and loads; and They are time sections node The active power consumed by photovoltaic power and loads; Time section The active power transmitted from the main grid to the distribution network; and These are the minimum and maximum allowable node voltage amplitudes for the distribution network, respectively. This represents the maximum allowable branch current amplitude in the distribution network.
8. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The low-carbon objective function in step 3 is: ； ； ； ； in, For long-term carbon emissions from power distribution networks; Carbon emissions resulting from electricity consumption in the power distribution network; Carbon emissions resulting from energy loss due to intelligent soft switching in hydrogen energy storage; This refers to the reduction in carbon emissions resulting from photovoltaic power generation. To calculate the total number of time sections; This represents the total number of time segments within the scheduling cycle. Carbon emission factor for electricity; For distribution network nodes; A collection of distribution network lines; For the line The resistance; Time section line Current on; A set of nodes that connect to intelligent soft switches in the distribution network; A collection of energy storage devices; Time section Energy storage devices Energy loss generated; A collection of photovoltaic access nodes; Time cross section node The amount of active power consumed by photovoltaic power generation.
9. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, The efficiency rolling mapping method in step 4 includes the following steps: Step 4.1.1: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switch containing hydrogen energy storage, combined with the annual natural day-level predicted source-load data, the average energy efficiency sequence of proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in Step 1, The daily offline scheduling plan for hydrogen storage was calculated recently. ; Step 4.1.2: Based on the established long-term low-carbon operation model of the distribution network with intelligent soft switch containing hydrogen energy storage, combined with the intraday hourly predicted source-load data, the average energy efficiency sequence of proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in Step 1, The daily hydrogen storage offline scheduling plan generated in step 4.1.1 Under constraints, the offline power control strategy for the hydrogen unit at the hourly level within the day was calculated; Step 4.1.3: Based on the function mapping relationship obtained from the intelligent soft-switching model of hydrogen energy storage considering the variable efficiency characteristics of hydrogen devices. The intraday hourly offline power control strategy for hydrogen devices generated in step 4.1.2 is mapped to the average energy efficiency sequences of the proton exchange membrane hydrogen electrolyzer and proton exchange membrane hydrogen fuel cell input in step 1. In this process, the efficiency pre-mapped sequence is obtained. ; Step 4.1.4: Update the efficiency pre-mapping sequence in real time based on the actual power and efficiency conditions of the hydrogen unit during operation. This reflects the actual operating conditions, resulting in an efficiency rolling mapping matrix, which is then applied to the real-time scheduling and control process. The efficiency rolling mapping matrix is expressed as follows: ； in, For the first Intra-day efficiency rolling mapping matrix; For the first The efficiency pre-mapping sequence for the first hour of the day; For the first The efficiency pre-mapped sequence for the second hour of the day; For the first The efficiency pre-mapped sequence for the 24th hour of the day; For the first The efficiency rolling mapping sequence for the first hour of the day; For the first The efficiency rolling mapping sequence for the second hour of the day.
10. The intelligent soft-switching energy dispatch method for hydrogen energy storage based on variable efficiency modeling according to claim 1, characterized in that, Step 4, the scheduling solution based on the two-level model predictive control method using SOC cut set decomposition, includes the following steps: Step 4.2.1: Set the scheduling period. The upper-level model's predictive control time granularity is the calendar day, and the lower-level model's predictive control time granularity is the hour. Set the scheduling control time to [time value missing]. ,in, To schedule the number of days at a given time point in real time, To set the hour for real-time scheduling, the initial time for operation and scheduling is set. , ; Step 4.2.2: Based on the input annual daily-level predicted source-load data and the established long-term low-carbon operation model of the distribution network with hydrogen energy storage intelligent soft switching, call the solver to generate an annual hydrogen energy storage scheduling plan with daily-level time granularity, as the first... TianSOC cut set ,in, For the first Net hydrogen inventory per day For the first The mass of hydrogen produced by the proton exchange membrane hydrogen electrolyzer; For the first The mass of hydrogen consumed by a proton exchange membrane hydrogen fuel cell. For the first Heavenly Net hydrogen intake per hour The interval is a natural day time segment. The time interval is the hourly cross-section. This represents the number of hours within a day, with a value of 24, and is continuously updated and calculated based on the scheduling process at the natural day time granularity. Step 4.2.3: Based on step 4.2.2, the generated first... TianSOC cut set Combining the input intraday hourly predicted source-load data, the established long-term low-carbon operation model of the distribution network with hydrogen-containing energy storage intelligent soft switch, and the efficiency rolling mapping matrix The solver is invoked to generate the daily power output of a proton exchange membrane hydrogen electrolyzer and a proton exchange membrane hydrogen fuel cell with hourly time granularity. , Control strategies are implemented and updated on a rolling basis at the hourly time granularity within the day. Step 4.2.4: Determine whether the daily scheduling plan for all time periods has been generated. If not, update... for ； Then return to step 4.2.3; if so, update the daily distribution network carbon emissions and photovoltaic absorption rate data, update the high-pressure gaseous hydrogen storage tank energy storage data, and update... Proceed to the next step; Step 4.2.5: Determine whether the generation of scheduling control plans for all time periods of the scheduling cycle has been completed. If not, update... for ； Then return to step 4.2.2; if so, end.