Hydrogen-doped comprehensive energy system scheduling method and system considering dynamic gas flow delay of pipe network

Abstract

The invention discloses a hydrogen-doped comprehensive energy system scheduling method and system considering pipe network dynamic airflow delay, and belongs to the technical field of electric power system operation control and comprehensive energy management. Space-time transmission distribution of hydrogen in the pipe network is predicted, the hydrogen energy demand urgency degree of the power grid side is quantified, and meanwhile discrete hydrogen flow in the pipe transmission process is dynamically aggregated into virtual schedulable resources; and matching urgent demands with virtual resources, searching a cooperative path meeting constraints in a space-time grid, screening an optimal path through multi-objective optimization, and finally decoding the optimal path into a specific pipe network scheduling instruction. According to the invention, through rolling prediction based on fluid space-time dynamic deduction and virtual aggregation mapping of management and storage resources, modeling is carried out for pipe network transmission characteristics, and space-time matching and collaborative optimization control of heterogeneous energy transmission delay and power grid instantaneous regulation requirements can be realized.

Description

Technical Field

[0001] This invention relates to power system operation control and integrated energy management technology, specifically to a hydrogen-infused integrated energy system scheduling method and system that considers the dynamic airflow delay in the pipeline network. Background Technology

[0002] Integrated energy systems have become an important means of improving renewable energy absorption and enhancing grid resilience in recent years. Among them, the electro-gas coupling system utilizes power-to-gas technology to convert unstable wind and solar power into hydrogen, which is then mixed and transmitted through pipelines. This hydrogen is then fed back to the grid at load centers via gas turbines or fuel cells, participating in grid regulation through a spatiotemporal energy transfer mechanism. Current dispatch control units typically monitor data at both the source and load ends through SCADA systems, aiming to support and optimize wide-area power balance through energy cascade utilization and multi-energy complementarity among various energy subsystems.

[0003] In existing technologies, scheduling strategies for hydrogen-infused integrated energy systems often employ time-sharing optimization algorithms based on gas network steady-state models or simple energy hub node models. These technical solutions typically assume that gas flow and composition are instantaneously balanced throughout the entire pipeline network within a single scheduling period, and assume that the energy injected upstream can be utilized downstream within the current period. Furthermore, they do not deeply distinguish the independent characteristics of hydrogen as a specific component in complex pipeline networks, including its mixing and movement, concentration evolution, and long-distance physical transport. They often simplify it into a general fluid transport problem controlled by flow constraints.

[0004] However, existing technical solutions based on quasi-steady state or large time-scale averaging ignore the significant transmission hysteresis, diffusion delay, and inertial dynamic processes of fluids, especially mixed gases, in long-distance pipelines. Since the physical transmission speed of gas is much lower than that of electricity, there is a huge time misalignment between the issuance of dispatching instructions and the actual energy response. Existing steady-state models cannot accurately quantify the actual lag time of the gas flow from the injection point to the regulating equipment and the actual concentration upon arrival. This can easily lead to regulation failures due to the failure to deliver hydrogen fuel during peak hours when the power grid urgently needs power support, or to safety and stability defects such as pipeline pressure exceeding limits or concentration exceeding standards due to advance regulation instructions. Summary of the Invention

[0005] Purpose of the invention: The purpose of this invention is to provide a method and system for scheduling a hydrogen-infused integrated energy system that considers the dynamic airflow delay in the pipeline network. By using rolling prediction based on fluid spatiotemporal dynamic deduction and virtual aggregation mapping of pipeline storage resources, the transmission characteristics of the pipeline network are modeled, which can realize the spatiotemporal matching and collaborative optimization control of heterogeneous energy transmission delay and instantaneous adjustment demand of the power grid.

[0006] Technical solution: The present invention provides a method for scheduling a hydrogen-doped integrated energy system that considers dynamic gas flow delay in a pipeline network, comprising:

[0007] Real-time operational data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network are obtained to form a real-time pipeline network status dataset;

[0008] Based on the real-time status dataset of the pipeline network, a pre-trained pipeline network dynamic simulation model or data-driven model is invoked to perform rolling predictions starting from the current moment, generating a spatiotemporal grid for hydrogen energy transmission.

[0009] Real-time acquisition of power grid frequency deviation data and node power deficit prediction data yields a real-time power grid dispatch demand dataset.

[0010] Obtain the performance parameters of hydrogen-using units at load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantify the urgency of hydrogen energy demand.

[0011] Based on the connectivity status of the hydrogen energy transmission spatiotemporal grid, the current hydrogen storage and corresponding airflow delay information of all hydrogen in the transmission process within the physical pipeline are statistically analyzed. The discretely distributed hydrogen flow segments are dynamically clustered to construct a virtual hydrogen energy pool state vector that can be regarded as a scheduling resource.

[0012] Using the virtual hydrogen energy reservoir state vector as resource points and the hydrogen-using units corresponding to the load points as target points, a spatiotemporal matching search is performed in the direction that satisfies the intensity gradient of hydrogen energy demand, based on the spatiotemporal grid of hydrogen energy transmission, and a candidate set of hydrogen energy flow collaborative paths is obtained.

[0013] With energy cost and grid power balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen blending safety threshold of the gas turbine unit as constraints, an objective function is established, and the final hydrogen energy flow collaborative path is obtained by screening from the candidate set of hydrogen energy flow collaborative paths based on the objective function.

[0014] Based on the final hydrogen energy flow collaborative path, a scheduling trajectory matrix is ​​constructed, and pipeline scheduling control instructions are generated by decoding the timing control nodes in the scheduling trajectory matrix.

[0015] Furthermore, the acquisition of real-time operational data from gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network forms a real-time pipeline network status dataset, including:

[0016] The hydrogen concentration and flow rate at the gas source point of the hydrogen-blended integrated energy system pipeline network are obtained, and the predicted inlet hydrogen concentration at each downstream node of the pipeline network is calculated by combining the diversion ratio of the hydrogen-blended integrated energy system pipeline network distribution point.

[0017] Real-time pipeline pressure data and cross-sectional flow data of upstream and downstream adjacent nodes of the hydrogen-blended integrated energy system network are collected simultaneously to calculate the flow-pressure gradient vector characterizing the airflow direction.

[0018] By combining the predicted inlet hydrogen concentration with the flow-pressure gradient vector, the hydrogen concentration of each node at the next sampling time is calculated, generating node hydrogen concentration parameters that include the spatiotemporal evolution of concentration.

[0019] The pipeline pressure data, the cross-sectional flow data, and the node hydrogen concentration parameters are aligned in time sequence and associated with the corresponding topological locations in the hydrogen-blended integrated energy system pipeline network, and integrated into a real-time pipeline network status dataset.

[0020] Furthermore, the step of generating a hydrogen energy transmission spatiotemporal grid by calling a pre-trained pipeline dynamic simulation model or data-driven model based on the real-time status dataset of the pipeline network and performing rolling predictions starting from the current moment includes:

[0021] Based on the pipeline pressure data and cross-sectional flow data in the real-time status dataset of the pipeline network, the initial steady-state power flow distribution of the pipeline network at the current moment is calculated using the preset gas pipeline network fluid dynamics equation.

[0022] Using the node hydrogen concentration parameters in the real-time status dataset of the pipeline network as boundary conditions, the convection and diffusion process of hydrogen under the initial steady-state power flow distribution of the pipeline network is simulated to generate a preliminary hydrogen energy spatiotemporal distribution grid.

[0023] Based on the cross-sectional flow data, the preliminary hydrogen energy spatiotemporal distribution grid is dynamically and continuously corrected, and the expected hydrogen arrival concentration and expected arrival time of each grid cell in the preliminary hydrogen energy spatiotemporal distribution grid are updated to obtain the hydrogen energy transmission spatiotemporal grid.

[0024] Furthermore, the real-time acquisition of power grid frequency deviation data and node power deficit prediction data yields a real-time power grid dispatch demand dataset, including:

[0025] The actual frequency of each key node in the power grid is monitored in real time, and the deviation between the actual frequency and the preset rated frequency is calculated to obtain power grid frequency deviation data.

[0026] Acquire renewable energy output forecast data and load forecast data, calculate the power generation and consumption difference of each key node in the power grid within the future dispatch cycle, and generate power deficit forecast data for key nodes.

[0027] By associating the key nodes in the power grid frequency deviation data and the key node power deficit prediction data that have overlapping geographical locations, and performing weighted fusion, a real-time power grid dispatch demand dataset that characterizes the intensity and location of power grid regulation demand is formed.

[0028] Furthermore, the process of obtaining the performance parameters of hydrogen-consuming units at load points corresponding to the real-time grid dispatch demand dataset, mapping the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantifying the urgency of hydrogen energy demand includes:

[0029] Obtain the type of hydrogen-using unit connected to the load point and its power-to-hydrogen-consumption conversion coefficient;

[0030] Based on the real-time grid dispatch demand dataset, a direct mapping relationship between grid power demand and hydrogen demand is established by combining the power-hydrogen consumption conversion coefficient. The equivalent hydrogen demand required by each load point to balance grid power at different times in the future is calculated through the direct mapping relationship between grid power demand and hydrogen demand.

[0031] On the spatiotemporal grid of hydrogen energy transmission, the grid nodes corresponding to each load point are located, and the hydrogen supply gap intensity under time constraints is calculated based on the equivalent hydrogen demand and the expected hydrogen arrival concentration and expected arrival time of the grid nodes corresponding to each load point.

[0032] The intensity of the hydrogen supply gap at all load points is normalized and sorted to generate a hydrogen demand urgency rating that characterizes the urgency and magnitude of demand at each grid node.

[0033] Furthermore, based on the connectivity state of the hydrogen energy transmission spatiotemporal grid, the current time's hydrogen inventory and corresponding gas flow delay information for all hydrogen gas segments in the transmission process within the physical pipeline are statistically analyzed. This is followed by dynamic clustering of the discretely distributed hydrogen gas flow segments to construct a virtual hydrogen reservoir state vector that can be viewed as a scheduling resource.

[0034] Analyze the spatiotemporal grid of hydrogen energy transmission to identify all hydrogen flow segments that have been sent from the gas source or storage point at the current moment but have not yet arrived at the target load point;

[0035] Extract the current spatial location, remaining transport distance, current airflow speed, and hydrogen inventory of each hydrogen flow segment in transit, calculate the estimated time to reach the target load point, and obtain airflow delay information;

[0036] Based on the gas flow delay information and the adjacency relationship of the flow segments in the pipeline network topology of the hydrogen-infused integrated energy system, multiple discrete hydrogen flow segments with similar expected arrival times and located in continuous pipeline segments are aggregated to form multiple virtual hydrogen resource clusters.

[0037] Each virtual hydrogen resource cluster is assigned attribute information, which together constitute a virtual hydrogen energy pool state vector. The attribute information includes the total hydrogen storage within the cluster, the current equivalent concentration location, and the available time window.

[0038] Furthermore, the virtual hydrogen reservoir state vector is used as a resource point, and the hydrogen-using unit corresponding to the load point is used as the target point. Based on the hydrogen energy transmission spatiotemporal grid, a spatiotemporal matching search is performed in the direction that satisfies the intensity gradient of hydrogen energy demand urgency, and a candidate set of hydrogen energy flow cooperative paths is obtained, including:

[0039] Using each virtual hydrogen resource cluster in the virtual hydrogen energy reservoir state vector as a resource point and the load point pointed to by the hydrogen energy demand urgency as the target load point, a breadth-first search is performed in the hydrogen energy transmission spatiotemporal grid to obtain a set of alternative transmission paths.

[0040] From the set of alternative transmission paths, select the path whose expected hydrogen concentration at each grid node meets the concentration threshold of the target load point, and accumulate the total transmission delay of the path to obtain the set of feasible paths.

[0041] From the set of feasible paths, feasible paths with a total transmission delay less than the preset demand time window are selected and bound to the corresponding virtual hydrogen resource clusters and target load points to form a candidate set of hydrogen energy flow collaborative paths.

[0042] Furthermore, the objective function is established with energy cost and grid power balance as targets, and the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen blending safety threshold of the gas turbine unit as constraints. Based on this objective function, the final hydrogen energy flow collaborative path is obtained by screening from the candidate set of hydrogen energy flow collaborative paths, including:

[0043] Establish a cost calculation model that includes hydrogen production costs, pipeline transportation costs, and regulation and compensation costs, as the target item for energy costs;

[0044] A model is established to minimize the sum of real-time power deficit data at each key node of the power grid, which serves as the target term for the power grid power balance.

[0045] Combining the energy cost target and the power grid balance target, the contribution value of the energy cost and the power grid balance of the hydrogen-using unit corresponding to each path in the candidate set of hydrogen energy flow collaborative paths is calculated.

[0046] The allowable ramp rate of the hydrogen-using unit electrolyzer is used as the time-series variation constraint, and the allowable hydrogen blending ratio of the hydrogen-using unit gas turbine is used as the concentration constraint.

[0047] Based on the time-series variation constraints and the concentration constraints, a multi-objective optimization algorithm is used to weigh and solve the contribution value, and the Pareto optimal solution is selected from the candidate set of hydrogen energy flow collaborative paths to determine the final hydrogen energy flow collaborative path.

[0048] Furthermore, the step of constructing a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decoding and generating pipeline scheduling control instructions from the timing control nodes in the scheduling trajectory matrix, includes:

[0049] The final hydrogen energy flow coordination path is unfolded according to the time sequence, and a scheduling instruction group containing instruction time, action type and operation amount is generated for each gas source point, storage point, distribution point and load point involved in the final hydrogen energy flow coordination path. The action type includes pressurization, diversion, injection and consumption.

[0050] All scheduling instruction groups are arranged in chronological order and associated with the topology nodes of the hydrogen-blended integrated energy system pipeline network to construct a two-dimensional scheduling trajectory matrix. The rows of the scheduling trajectory matrix represent the topology nodes of the hydrogen-blended integrated energy system pipeline network, and the columns represent the time sequence.

[0051] Extract the timing control nodes and corresponding control instructions corresponding to the scheduling instruction groups to be executed in the current scheduling period from the scheduling trajectory matrix;

[0052] The timing control node and the control command are encoded into a standard communication protocol format and sent to the topology node corresponding to the hydrogen-blended integrated energy system pipeline network to generate pipeline network scheduling control commands.

[0053] Based on the same inventive concept, the present invention provides a hydrogen-doped integrated energy system scheduling system that considers dynamic gas flow delay in pipeline networks, comprising:

[0054] The data acquisition module is used to acquire real-time operating data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network, forming a real-time status dataset of the pipeline network;

[0055] The grid generation module is used to generate a spatiotemporal grid for hydrogen energy transmission by calling a pre-trained dynamic simulation model or data-driven model of the pipeline network based on the real-time status dataset of the pipeline network, and performing rolling prediction starting from the current moment.

[0056] The real-time dispatch demand dataset generation module is used to collect power grid frequency deviation data and node power deficit prediction data in real time to obtain the real-time dispatch demand dataset of the power grid.

[0057] The hydrogen energy demand urgency generation module is used to obtain the performance parameters of hydrogen-using units at load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantify the hydrogen energy demand urgency.

[0058] The state vector construction module is used to statistically analyze the hydrogen storage and corresponding gas flow delay information of all hydrogens in the current transmission process in the physical pipeline network based on the connectivity state of the hydrogen energy transmission spatiotemporal grid, dynamically cluster the discrete hydrogen flow segments, and construct a virtual hydrogen energy pool state vector that can be regarded as a scheduling resource.

[0059] The hydrogen energy flow collaborative path candidate set arrangement module is used to take the virtual hydrogen energy reservoir state vector as the resource point and the hydrogen-using unit corresponding to the load point as the target point, and perform spatiotemporal matching search in the direction that satisfies the hydrogen energy demand urgency intensity gradient according to the hydrogen energy transmission spatiotemporal grid, and arrange to obtain the hydrogen energy flow collaborative path candidate set.

[0060] The hydrogen energy flow collaborative path determination module is used to establish an objective function with energy cost and grid power balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen doping safety threshold of the gas turbine unit as constraints. Based on the objective function, the module selects the final hydrogen energy flow collaborative path from the candidate set of hydrogen energy flow collaborative paths.

[0061] The control command generation module is used to construct a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decode and generate pipeline scheduling control commands from the timing control nodes in the scheduling trajectory matrix.

[0062] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are as follows:

[0063] (1) By constructing a dynamic rolling simulation model of the gas network based on physical fluid dynamics and concentration diffusion mechanism, this invention can generate a quantitative spatiotemporal grid of hydrogen energy transmission in real time, which includes spatial location, airflow velocity and concentration distribution. Thus, when the power grid initiates a dispatch request, it can predict the time and concentration of hydrogen gas masses at different locations in the pipeline to reach the load point, avoiding the actual energy supply and demand mismatch caused by ignoring gas transmission delay, so that the hydrogen energy resources based on pipeline transmission can meet the power regulation needs of the power grid.

[0064] (2) By identifying and aggregating discrete hydrogen flow segments in the pipeline that are in the transmission state, the present invention realizes dynamic management of the originally passively transported gas and reconstructs it into a virtual hydrogen energy reservoir state vector with a clear time window and energy attributes. This mechanism fully taps the buffer and energy storage potential of the pipeline network itself without adding physical gas storage facilities, and provides an instant response resource pool for the renewable energy fluctuations faced by the power grid.

[0065] (3) By introducing dynamic airflow delay variables into the multi-objective optimization framework, this invention establishes a two-stage verification mechanism to ensure that each generated scheduling instruction is physically verified and temporally coupled with the grid demand; under the condition of fluctuating hydrogen doping ratio, it balances the safety boundary and regulation capability of electrical equipment such as gas turbine units; and through cost analysis and balance evaluation of the whole life cycle, it maximizes the comprehensive benefits under the coordinated operation of the power grid and gas grid. Attached Figure Description

[0066] Figure 1 This is a flowchart illustrating a hydrogen-doped integrated energy system scheduling method that considers dynamic airflow delay in pipelines, as disclosed in an embodiment of the present invention.

[0067] Figure 2 This is a schematic diagram of the spatiotemporal grid for hydrogen energy transmission disclosed in an embodiment of the present invention;

[0068] Figure 3 This is a schematic diagram of the pipeline topology of the hydrogen-doped integrated energy system disclosed in the embodiments of the present invention;

[0069] Figure 4 This is a schematic diagram of the structure of a hydrogen-doped integrated energy system scheduling system that considers the dynamic airflow delay of the pipeline network, as disclosed in an embodiment of the present invention. Detailed Implementation

[0070] The technical solution of the present invention will now be described in detail with reference to specific embodiments and accompanying drawings.

[0071] Example 1

[0072] like Figure 1 As shown, the present invention provides a hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in pipeline networks, comprising the following steps:

[0073] S1. Obtain real-time operating data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network to form a real-time status dataset of the pipeline network.

[0074] The specific implementation process of step S1 is as follows:

[0075] S1.1 Obtain the hydrogen concentration and flow rate at the gas source points of the hydrogen-blended integrated energy system pipeline network, and calculate the predicted inlet hydrogen concentration at each downstream node of the pipeline network based on the diversion ratio of the distribution points. Specifically:

[0076] Real-time operational data is collected starting from four key topological points in the hydrogen-blended integrated energy system pipeline network: gas source, storage, distribution, and load. Data from the gas source includes hydrogen concentration and flow rate; data from the storage point includes stored hydrogen concentration and injection / extraction flow rates (the storage point can be considered the origin); data from the distribution point includes the diversion ratio and inlet / outlet flow rates; and data from the load point includes real-time consumption flow rates. Based on the hydrogen concentration and flow rate of the gas source, and combined with the diversion ratio recorded at its downstream distribution points, the predicted inlet hydrogen concentration at each downstream node is calculated, based on the principle of mass conservation.

[0077] ;

[0078] in, This represents the predicted inlet hydrogen concentration for downstream nodes. It is a volume fraction between 0 and 1. Equal to the predicted value of inlet hydrogen concentration ; The gas supply or transmission flow rate of the upstream node, expressed in standard cubic meters per hour; This represents the hydrogen concentration at the upstream node corresponding to the downstream node. This represents the traffic splitting ratio from the upstream node corresponding to the downstream node to the current downstream node. A coefficient between 0 and 1; This represents the total intake flow rate of downstream nodes, expressed in standard cubic meters per hour. The setting is based on the general engineering principles of gas pipeline mass balance.

[0079] S1.2. Simultaneously collect real-time pipeline pressure data and cross-sectional flow data of upstream and downstream adjacent nodes in the hydrogen-infused integrated energy system pipeline network, and calculate the flow-pressure gradient vector characterizing the airflow direction. Specifically:

[0080] While performing step S1.1, real-time pipeline pressure data of all nodes in the hydrogen-blended integrated energy system pipeline network, as well as cross-sectional flow data between each node and all directly connected upstream and downstream nodes, are simultaneously collected. Based on the real-time pipeline pressure data and cross-sectional flow data, the flow-pressure gradient vector characterizing the airflow direction is calculated. This vector comprehensively reflects the spatial variation trend of flow and pressure at this node.

[0081] S1.3. By fusing the predicted inlet hydrogen concentration with the flow-pressure gradient vector, the hydrogen concentration at each node at the next sampling time is calculated, generating node hydrogen concentration parameters that include the spatiotemporal evolution of concentration. Specifically:

[0082] The predicted inlet hydrogen concentration calculated in steps S1.1 and S1.2 is combined. The flow pressure gradient vector of the current node Using a convection-diffusion process model, the hydrogen concentration at the current node at the next sampling time is calculated, generating node hydrogen concentration parameters containing spatiotemporal evolution information of the concentration. The convection-diffusion process model is as follows:

[0083] ;

[0084] in, This represents the predicted hydrogen concentration at the node in the next time step. Let be the hydrogen concentration at the node at the current moment. The initial value comes from the sensor or the previous round of calculation; It is a dimensionless attenuation coefficient, the value of which is obtained by historical fitting of 200 sets of industrial sensor measured data, and is used to adjust the rate at which the concentration approaches the inlet value. The magnitude of the current node's flow pressure gradient vector; The gradient normalization factor is the maximum gradient value recorded during the long-term operation of the system, such that the ratio... It is dimensionless and ranges from 0 to 1.

[0085] S1.4 Align pipeline pressure data, cross-sectional flow data, and node hydrogen concentration parameters according to time sequence and associate them with the corresponding topological locations in the hydrogen-blended integrated energy system pipeline network to integrate them into a real-time pipeline network status dataset.

[0086] All collected pipeline pressure data, cross-sectional flow data, and calculated node hydrogen concentration parameters are time-aligned according to a unified timestamp and associated with their unique position codes in the hydrogen-blended integrated energy system pipeline network topology, thus forming a structured real-time pipeline network status dataset.

[0087] For example, the process of assembling the real-time status dataset of the pipeline network in step S1 is as follows: Assume there is a gas source point S1 in the pipeline network with a hydrogen concentration of 0.95 and a gas flow rate of 500 standard cubic meters per hour. Downstream, there is a distribution point J1, whose flow diversion ratio from the gas source point S1 is 0.6. According to the formula, the predicted hydrogen concentration at the inlet of the downstream pipeline section flowing towards the distribution point J1 is calculated. The value is 0.95 because there is only one upstream source. The real-time pipeline pressure at distribution point J1 was synchronously acquired to be 2.1 MPa, with an upstream cross-sectional flow rate of 500 standard cubic meters per hour and two downstream cross-sectional flow rates of 300 and 200 standard cubic meters per hour, respectively. Based on these flow rate differences and pressure data, the flow-pressure gradient vector at distribution point J1 was calculated. Its size The calculated value is 15, a dimensionless value after unit standardization. Let the normalization factor be... The attenuation coefficient is 100. The value is 0.1. If the hydrogen concentration at node J1 at the current moment... The predicted inlet hydrogen concentration is 0.93. If the value is 0.95, then the node hydrogen concentration parameter at the next sampling time can be calculated according to the formula. The pipeline pressure of distribution point J1 (2.1 MPa), upstream flow rate of 500 standard cubic meters per hour, downstream flow rates of 300 and 200 standard cubic meters per hour, and the node hydrogen concentration parameter of 0.9303, along with the timestamp and topological location code J1, are written into a single record in the pipeline network real-time status dataset. Data from all other nodes in the pipeline network are collected, calculated, and integrated synchronously using the same method to form a complete pipeline network real-time status dataset.

[0088] S2. Based on the real-time status dataset of the pipeline network, call the pre-trained dynamic simulation model or data-driven model of the pipeline network, and perform rolling prediction starting from the current moment to generate a spatiotemporal grid for hydrogen energy transmission.

[0089] like Figure 2 As shown, the specific implementation process of step S2 is as follows:

[0090] S2.1. Based on the pipeline pressure and cross-sectional flow data in the real-time pipeline network status dataset, the initial steady-state power flow distribution of the pipeline network at the current moment is calculated using the preset gas pipeline network fluid dynamics equations. Specifically:

[0091] Based on the real-time pipeline status dataset obtained in step S1, a pipeline dynamic simulation model pre-trained using historical operating data is invoked. This model, starting from the current moment, predicts changes in the pipeline status over a future period through iterative calculations. The first step in generating the spatiotemporal grid for hydrogen energy transmission is to utilize the pipeline pressure and cross-sectional flow data from the real-time pipeline status dataset to construct the initial steady-state power flow distribution of the entire pipeline network at the current moment, using a pre-defined gas pipeline fluid dynamics equation to describe the gas flow pattern within the pipeline. The core of the gas pipeline fluid dynamics equation is the conservation of mass and momentum. For each pipeline segment, the relationship between its flow rate and the pressure at both ends is as follows:

[0092] ;

[0093] in, For the node Flow to Node The cross-sectional flow rate is expressed in standard cubic meters per second. For pipe section The flow capacity coefficient is a constant that takes into account the pipe diameter, length, roughness and gas characteristics. Its value is obtained by fitting and calibrating the pipe design parameters with 100 sets of historical steady-state operation data, and the unit is standard cubic meters per second per Pa. and They are nodes and nodes Absolute pressure, measured in Pascals; function Used to determine the direction of airflow. The design is based on standard engineering methods for steady-state hydraulic calculations of gas pipeline networks. The equations for all pipe sections are combined with the flow balance equations at the nodes: Solving this equation yields the initial steady-state power flow distribution of the pipeline network at the current moment, which satisfies all boundary conditions, i.e., the flow rate of each pipeline and the pressure of each node.

[0094] S2.2 Using the node hydrogen concentration parameters in the real-time pipeline network data set as boundary conditions, simulate the convection and diffusion process of hydrogen under the initial steady-state power flow distribution of the pipeline network, generating a preliminary spatiotemporal distribution grid for hydrogen energy. Details are as follows:

[0095] The node hydrogen concentration parameters are calculated from the real-time status data of the pipeline network. As the initial concentration boundary condition, it is applied to the calculated initial steady-state tidal flow distribution. The propagation process of hydrogen under this fixed flow field is simulated, considering both convection caused by airflow and diffusion due to concentration differences. The pipeline network is spatiotemporally discretized, dividing the pipeline into several length units and the prediction period into several time steps, thus forming a preliminary spatiotemporal distribution grid of hydrogen energy composed of spatial grid points and time layers. Each grid unit records the predicted hydrogen concentration at its location at the corresponding time, and its update calculation is based on an explicit difference scheme of the one-dimensional convection-diffusion equation:

[0096] ;

[0097] ;

[0098] ;

[0099] in, In spatial location ,time The predicted hydrogen concentration of the grid cells; For spatial location The airflow velocity inside the duct is determined by its spatial position. Flow rate of the pipe section Calculated by dividing by the cross-sectional area of ​​the pipe, in meters per second; For time step; This is the spatial step size, in meters. is the effective diffusion coefficient of hydrogen, expressed in square meters per second, and its value is determined based on experimental data of gas properties and pipe flow turbulence intensity. In spatial location ,time The predicted hydrogen concentration of the grid cells; In spatial location ,time The predicted hydrogen concentration of the grid cells; In spatial location ,time The predicted hydrogen concentration of the grid cells; This refers to the change in concentration caused by convection, specifically the amount of hydrogen carried away or brought in by the physical behavior of gas flow. In other words, it represents the change in concentration at a specific point (e.g., (x, t)) caused by hydrogen molecules moving along the pipeline at the speed of the natural gas flow. The change in concentration due to diffusion is the concentration change caused by the physical behavior of molecular diffusion and turbulent mixing. Even without overall gas flow velocity, hydrogen will automatically diffuse and permeate from high concentration areas to low concentration areas due to the existence of a concentration gradient.

[0100] S2.3. Based on the cross-sectional flow data, the preliminary hydrogen energy spatiotemporal distribution grid is dynamically and continuously corrected to update the expected hydrogen arrival concentration and expected arrival time of each grid cell in the preliminary hydrogen energy spatiotemporal distribution grid, thereby obtaining the hydrogen energy transmission spatiotemporal grid.

[0101] At each new sampling time, the spatial location in the model is updated with the latest cross-sectional flow data. airflow velocity inside the pipe Using the concentration distribution calculated by the grid at that moment as the new initial condition, the aforementioned convection-diffusion simulation is re-executed, thereby continuously updating the expected hydrogen arrival concentration and expected arrival time of each grid cell in the initial hydrogen energy spatiotemporal distribution grid. Finally, the grid structure, after dynamic rolling correction and containing the predicted hydrogen concentration and gas flow arrival time information of each pipeline node at future moments, is output as the hydrogen energy transport spatiotemporal grid.

[0102] For example, in step S2, the process of generating the spatiotemporal grid for hydrogen energy transmission is as follows: Read the current time data from the real-time status dataset of the pipeline network, for example, the pressure at node A is 3.0 MPa, the pressure at node B is 2.8 MPa, and the flow capacity coefficient of pipe segment AB. The value is 0.05 standard cubic meters per second per Pa. Calculated based on the steady-state flow rate formula. The flow rate is standard cubic meters per second, flowing from A to B in the forward direction. This iterative calculation yields the initial steady-state power flow distribution for the entire network. Assume the hydrogen concentration parameter at node A is 0.90 and at node B it is 0.88. Discretize the pipe segment AB, setting the spatial step size... 1000 meters, time step The time is 10 seconds. The airflow velocity at the starting point of pipe section AB is taken. At 5 meters per second, the effective diffusion coefficient is... The concentration is 0.1 square meters per second. Based on the convection-diffusion equation, the predicted hydrogen concentration in a grid cell 1000 meters from the starting point can be calculated after 10 seconds. For example, substituting into the formula yields... .

[0103] At the next sampling time, the cross-sectional flow data is updated, and the new airflow velocity of pipe segment AB is calculated. If the velocity is 5.2 meters per second, the concentration distribution of the entire network calculated at the previous moment is used as the initial value, and the simulation is repeated to update the expected hydrogen arrival concentration and expected arrival time of each grid cell. After multiple time steps of rolling prediction and correction, a spatiotemporal grid for hydrogen energy transport is finally formed, in which each grid node records information such as "Node B, expected hydrogen concentration 0.878 in the next 300 seconds, expected gas flow arrival time from A 280 seconds".

[0104] S3. Real-time acquisition of power grid frequency deviation data and node power deficit prediction data to obtain a real-time power grid dispatch demand dataset.

[0105] The specific implementation process of step S3 is as follows:

[0106] S3.1 Real-time monitoring of the actual frequency of each key node in the power grid, calculation of the deviation between the actual frequency and the preset rated frequency, and obtaining power grid frequency deviation data;

[0107] S3.2 Obtain renewable energy output forecast data and load forecast data, calculate the power generation and consumption difference of each key node in the power grid within the future dispatch cycle, and generate power deficit forecast data of key nodes.

[0108] S3.3. Collect key nodes whose geographical locations overlap with the associated grid frequency deviation data and the key node power deficit prediction data, and perform weighted fusion to form a real-time grid dispatch demand dataset that characterizes the intensity and location of grid regulation demand.

[0109] Specifically, the real-time acquisition process begins with the synchronous monitoring of the actual frequencies of all nodes in the power grid marked as critical nodes. These critical nodes are typically power grid hubs or sites with significant regulation capabilities. Actual frequency values ​​are acquired at a sampling rate of at least once per second using frequency measuring devices installed at the nodes. The difference between this actual frequency and the preset rated frequency of the power grid is calculated to obtain the power grid frequency deviation data for that node, i.e.:

[0110] ;

[0111] in, The preset rated frequency for the power grid is set according to the State Grid standard, for example, a value of 50.00 Hz; As a key node The actual frequency obtained from the above measurement, in Hertz; As a key node The frequency deviation is measured in Hertz (Hz), with positive values ​​indicating a higher frequency and negative values ​​indicating a lower frequency. Simultaneously, renewable energy output forecast data and load forecast data for the next scheduling cycle are obtained from the grid energy management system. The renewable energy output forecast data includes the projected power generation from wind farms, photovoltaic power plants, etc., while the load forecast data includes the projected power consumption at each node. For each key node... Calculate its future sampling times The difference between power generation and power consumption is used to generate nodal power deficit prediction data:

[0112] ;

[0113] in, As a key node At any moment The predicted total power generation, It is the sum of the projected output of renewable energy and the planned output of conventional units; As a key node At any moment Predicted load power; As a key node At any moment The predicted power deficit is expressed in megawatts (MW), with positive values ​​indicating a power surplus and negative values ​​indicating a power shortage. The data is based on a power balance prediction model commonly used in power grid dispatching.

[0114] Nodes whose geographical locations overlap in the correlated power grid frequency deviation data and the nodal power deficit prediction data, i.e., targeting the same critical node. To make its key nodes frequency deviation Power deficit forecasts for the near future To integrate, among which, This is the nearest predicted time after the current time. Since the physical meaning and dimensions of these two values ​​differ, normalization is required first. For frequency deviation, the absolute value of the maximum allowable frequency deviation of the power grid is used. Normalization is applied; this value is set based on power grid safety operation procedures, for example, 0.2 Hz. For power deficits, critical nodes are used. Maximum transmission capacity The value is normalized and determined based on the line and transformer capacity. The normalized frequency deviation index is:

[0115] ;

[0116] The power deficit index is:

[0117] ;

[0118] Then, the key nodes are calculated using a weighted fusion formula. The intensity of power grid regulation demand :

[0119] ;

[0120] in, and These are the weighting coefficients for frequency deviation and power deficit, respectively, satisfying... Its setting is based on the power grid dispatch strategy, such as when more emphasis is placed on instantaneous frequency stability. Take 0.6, Take 0.4; As a key node The grid regulation demand intensity is a dimensionless value between 0 and 1, with a larger value indicating a more urgent need for regulation resources at that node. The grid regulation demand intensity of all key nodes, along with their node numbers and geographical location information, are organized in a time series to form a real-time grid dispatch demand dataset.

[0121] For example, assume the power grid has a preset rated frequency. The actual frequency of key node A was detected at 50.0 Hz. If the frequency is 49.8 Hz, then calculate the frequency deviation of critical node A. Hertz. Renewable energy output forecast data for the next 5 minutes at this node is obtained from the energy management system; the predicted total power generation of critical node A. The predicted load power of critical node A is 80 MW. Calculate the predicted power deficit for critical node A, given a capacity of 100 MW. Megawatts. Sets the maximum permissible absolute value of the grid's frequency deviation. The maximum transmission capacity of critical node A is 0.2 Hz. 100 megawatts. Normalized frequency deviation index. Power deficit index Set the weighting coefficient for frequency deviation. Weighting factor for power deficit Calculate the grid regulation demand intensity of key node A. The regulation demand intensity of critical node A (0.68), along with its node number A and geographical coordinates, is written into a record in the real-time power grid dispatch demand dataset. The same operation is performed on all other critical nodes in the power grid, thus forming a complete real-time power grid dispatch demand dataset.

[0122] S4. Obtain the performance parameters of hydrogen-using units at the load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the spatiotemporal grid of hydrogen energy transmission, and quantify the urgency of hydrogen energy demand.

[0123] The specific implementation process of step S4 is as follows:

[0124] S4.1 Obtain the type of hydrogen-using unit connected to the load point and its power-to-hydrogen-consumption conversion coefficient.

[0125] From the load point configuration database of the hydrogen-blended integrated energy system pipeline network, obtain the type of hydrogen-consuming unit connected to each load point and its corresponding performance parameters. The most critical performance parameter is the power-to-hydrogen consumption conversion factor. This coefficient represents the volumetric hydrogen consumption under standard conditions corresponding to each unit of power generated or consumed by the hydrogen unit, expressed in standard cubic meters per megawatt-hour. For example, the power-to-hydrogen conversion coefficients for gas turbine units and fuel cells differ; their values ​​are provided in the equipment manufacturer's technical manual and verified based on field operating data.

[0126] S4.2. Based on the real-time power grid dispatch demand dataset and the power-hydrogen consumption conversion coefficient, establish a direct mapping relationship between power grid demand and hydrogen demand. Calculate the equivalent hydrogen demand required by each load point at different future times to balance the power grid through this direct mapping relationship. Details are as follows:

[0127] Based on the real-time power grid dispatch demand dataset obtained in step S3, especially for each key node... At any moment Power deficit prediction and each key node By combining the load point information associated with the hydrogen-blended integrated energy system network with the power-hydrogen consumption conversion coefficient, a direct mapping relationship between grid power demand and hydrogen demand is established. Based on this direct mapping relationship, the equivalent hydrogen demand required by each calculated load point to balance grid power at different future times is calculated.

[0128] ;

[0129] in, For key node i at time... The predicted power deficit is in megawatts. The power-hydrogen consumption conversion coefficient of the hydrogen unit connected to the load point associated with the critical node i in the hydrogen-blended integrated energy system pipeline network; For critical node i, the load point associated with the hydrogen-blended integrated energy system network at time t is... The equivalent hydrogen demand required to balance power output is expressed in standard cubic meters. This is based on the fundamental physical principles of energy conversion and the equipment's characteristic parameters.

[0130] S4.3. On the spatiotemporal grid of hydrogen energy transmission, locate the grid nodes corresponding to each load point, and calculate the hydrogen supply gap intensity under time constraints based on the equivalent hydrogen demand and the expected hydrogen arrival concentration and expected arrival time of the grid nodes corresponding to each load point. Specifically:

[0131] On the hydrogen energy transmission spatiotemporal grid generated in step S2, the grid nodes corresponding to each load point are located based on the pipeline topology. The predicted hydrogen arrival concentrations recorded by these grid nodes in the hydrogen energy transmission spatiotemporal grid are then read. Compared with the predicted arrival time (i.e., the time when hydrogen arrives at the grid node corresponding to the load point as predicted by the hydrogen energy transfer spatiotemporal grid), It needs to be determined that if the hydrogen pipeline network can provide sufficient concentration and flow rate of hydrogen at the required time, there is no shortage; otherwise, a supply shortage exists. The intensity of the hydrogen supply shortage. The calculation comprehensively considers the effects of hydrogen demand, supply time difference, and insufficient concentration. The formula is as follows:

[0132] ;

[0133] in, For time penalty items, ; For concentration penalty items, ; The moment when hydrogen is required to meet the grid demand, i.e. the moment when the power deficit occurs; The time when hydrogen arrives at the grid node corresponding to the load point, as predicted by the spatiotemporal grid for hydrogen energy transmission. The minimum hydrogen concentration threshold required for the normal operation of the hydrogen unit at the load point, as predicted by the spatiotemporal grid for hydrogen energy transmission, is determined by the equipment performance parameters. The predicted hydrogen concentration; function Ensure only when That is, the hydrogen arrives late, or That is, the multiplier is positive only when the concentration is insufficient; otherwise, it is zero, indicating that there is no deficiency in that aspect. This formula calculates... The unit is standard cubic meters, and the larger the value, the more severe the supply gap under the dual constraints of time and concentration.

[0134] S4.4. Normalize and sort the hydrogen supply gap intensity at all load points to generate a hydrogen demand urgency rating that characterizes the urgency and magnitude of demand at each grid node. Specifically:

[0135] Hydrogen supply gap intensity calculated for all load points at all scheduling times Normalization is performed to eliminate the influence of absolute dimensions. The normalization method uses the global maximum value method, i.e.:

[0136] ;

[0137] in, This represents the maximum value of the shortage intensity at all load points and all times within the current scheduling cycle. It also represents the normalized hydrogen supply shortage intensity. The hydrogen energy demand urgency is determined by sorting the values ​​from largest to smallest, along with their corresponding load point identifiers and demand time information, forming an ordered list. This list represents the urgency and magnitude of the demand at each node. The hydrogen energy demand urgency is a sorted list, where each item includes the load point, time, and normalized urgency value.

[0138] For example, suppose the critical node L1 is associated with a load point in the hydrogen-blended integrated energy system network. Connecting a gas turbine, the critical node L1 is the load point associated with the hydrogen-blended integrated energy system network. Power-to-hydrogen conversion coefficient The value is 250 standard cubic meters per megawatt-hour. According to the real-time power grid dispatch demand dataset, the critical node L1 has a predicted power deficit at time t=300 seconds. Megawatts (MW), with negative values ​​indicating power shortages. Therefore, the equivalent hydrogen demand required to balance power at time t=300 seconds is the load point associated with critical node L1 in the hydrogen-blended integrated energy system network. Standard cubic meters. Locating the load point on the spatiotemporal grid of hydrogen energy transmission. The corresponding grid node is used to read the predicted hydrogen concentration at t=300 seconds. The value is 0.88, which is the hydrogen arrival point predicted by the spatiotemporal grid for hydrogen energy transfer. The corresponding grid node's time The minimum hydrogen concentration threshold required for this gas turbine is 350 seconds. It is 0.85. The power grid requires hydrogen to be in If the time limit is reached in seconds, then a time penalty will apply. Concentration penalty item Therefore, the load point hydrogen supply gap intensity Standard cubic meters. Assuming the maximum shortfall intensity at all load points at all times within the current scheduling cycle is 5000 standard cubic meters, then after normalization... Load point Add the urgency value of 0.125 at time 300 seconds to the list. After performing the same calculation and normalization for all load points and all demand times, sort them in descending order of urgency value to obtain the final hydrogen energy demand urgency list, for example: [(load points (t=200, urgency 0.85), (load point) ,t=150, urgency 0.72),…,(load point) (t=300, urgency 0.125).

[0139] S5. Based on the connectivity status of the spatiotemporal grid for hydrogen energy transmission, statistically analyze the hydrogen inventory and corresponding gas flow delay information of all hydrogen gas in the current physical pipeline network that is in the process of transmission. Dynamically cluster the discretely distributed hydrogen gas flow segments to construct a virtual hydrogen energy reservoir state vector that can be regarded as a scheduling resource.

[0140] The specific implementation process of step S5 is as follows:

[0141] S5.1 Analyze the spatiotemporal grid of hydrogen energy transmission to identify all hydrogen flow segments currently en route that have departed from the gas source or storage point but have not yet arrived at the target load point. Specifically:

[0142] Step S5 constructs the virtual hydrogen reservoir state vector, which begins by analyzing the hydrogen transport spatiotemporal grid generated in step S2. The goal of this analysis is to identify all hydrogen units that have been emitted from the gas source or storage point at the current moment but have not yet arrived at any target load point according to grid predictions. These hydrogen units in transit are defined as hydrogen flow segments. Each hydrogen flow segment in transit is identified by its unique index in the spatiotemporal grid and carries a set of state attributes.

[0143] S5.2 Extract the current spatial location, remaining transport distance, current airflow speed, and hydrogen inventory for each hydrogen gas flow segment in transit, calculate the estimated arrival time at the target load point, and obtain the airflow delay information. Details are as follows:

[0144] Four core attributes are extracted from the spatiotemporal grid of hydrogen energy transmission for each en route hydrogen flow segment: current spatial location, defined by its location on the pipeline and its relative distance along the pipeline length; remaining transport distance, which is the total length of the pipeline from the current spatial location to its preset target load point along the shortest path of the pipeline network topology, in meters; current flow velocity, calculated by dividing the cross-sectional flow rate of the pipeline where the flow segment is located at the current moment by the cross-sectional area of ​​the pipeline, in meters per second; and hydrogen inventory, calculated based on the pipeline volume occupied by the flow segment, the current hydrogen concentration, and the gas density under standard conditions, in standard cubic meters.

[0145] Using the remaining transport distance and the current airflow velocity, the estimated time for this flow segment to reach the target load point can be calculated, i.e., the airflow delay information. :

[0146] ;

[0147] in, The remaining transportation distance; The current airflow velocity. The calculated airflow delay information. The unit is seconds.

[0148] S5.3. Based on the gas flow delay information and the adjacency relationship of the flow segments in the pipeline network topology of the hydrogen-blended integrated energy system, multiple discrete hydrogen flow segments with similar expected arrival times and located in continuous pipeline sections are aggregated to form multiple virtual hydrogen resource clusters. Specifically:

[0149] Dynamic clustering of discrete hydrogen flow segments is performed based on two criteria: similar flow delay information and continuous pipeline segments with consistent transmission directions within the hydrogen-blended integrated energy system network topology. A time-aggregation threshold is set for determining similar flow delays. For example, 300 seconds, this threshold is set based on scheduling cycle length and engineering experience. This applies when the estimated arrival time difference between two or more en route hydrogen gas flow segments... Less than If they are located between adjacent pipe nodes on the pipe network topology diagram and their airflow direction is continuous, then they are grouped into the same set. The pipe network topology diagram is as follows: Figure 3 As shown. This aggregation process is implemented through an iterative graph search algorithm. The algorithm starts from any unclustered flow segment and searches for other flow segments that satisfy the above spatiotemporal proximity conditions until no new flow segments can be added, thus forming a virtual hydrogen resource cluster.

[0150] S5.4 Assign attribute information to each virtual hydrogen resource cluster, which together constitute the virtual hydrogen energy pool state vector. The attribute information includes the total hydrogen inventory within the cluster, the current equivalent concentration location, and the available time window. Details are as follows:

[0151] Each virtual hydrogen resource cluster needs to be calculated and assigned three attributes, the first being the total hydrogen inventory within the cluster. :

[0152] ;

[0153] in, It is the set of all flow segments within the virtual hydrogen resource cluster; For the first The hydrogen stockpile in each flow segment. The second item is the current equivalent concentration location. The weighted average of existing assets is used for calculation:

[0154] ;

[0155] in, For the first The spatial coordinate vector of each flow segment. The third term is the available time window. ,in, , , and These are the earliest and latest estimated arrival times for the flow segments within the cluster, respectively.

[0156] Finally, the attribute information of all identified and constructed virtual hydrogen resource clusters at the current moment is arranged in order of cluster number to form the virtual hydrogen energy pool state vector. Each element It is a subvector that contains the first... Three attribute information of a cluster: , and .

[0157] For example, after parsing the spatiotemporal grid of hydrogen energy transport, three hydrogen gas flow segments F1, F2, and F3 were identified. Their attributes were extracted: F1 is currently located at the midpoint of pipeline P1, and the remaining transport distance of the hydrogen gas flow segment F1 is [not specified]. The current airflow velocity in the F1 section of the hydrogen flow path is 5000 meters. At 10 meters per second, hydrogen reserves Calculate the gas flow delay information for hydrogen gas flow segment F1 during the journey, given a volume of 200 standard cubic meters. Seconds. F2 is located at the connection point between the end of pipeline P1 and the beginning of pipeline P2. The remaining transport distance is 3000 meters, the current airflow speed is 10 meters per second, and the hydrogen supply is 150 standard cubic meters. Seconds. F3 is located on pipeline P2, with a remaining transport distance of 1000 meters, a current airflow speed of 10 meters per second, and a hydrogen supply of 100 standard cubic meters. Seconds. Set the time aggregation threshold. The time difference is 300 seconds. Examine the relationships between flow segments: F1 and F2 have a time difference of 200 seconds, less than 300 seconds, and they are topologically adjacent (P1's end connects to P2's starting point), flowing in the same direction towards the load point, therefore they can converge. F2 and F3 have a time difference of 200 seconds, less than 300 seconds, are topologically adjacent, and flow in the same direction, therefore they can converge. F1 and F3 have a time difference of 400 seconds, greater than 300 seconds, but since F1 converges with F2 and F2 converges with F3, based on transitivity, F1, F2, and F3 can be grouped into the same cluster. Calculate the attributes of this virtual hydrogen resource cluster: total hydrogen storage within the cluster. Standard cubic meters. For simplicity, assume spatial locations are represented by one-dimensional coordinates, with the coordinates of F1, F2, and F3 being respectively... rice, rice, If the value is in meters, then the current equivalent central location is... Meters. Earliest estimated arrival time of the flow segment within the cluster. Seconds, the latest estimated arrival time of the flow segment within the cluster. seconds, then Assuming there is another independent flow segment F4 in the system that does not meet the polymerization conditions, with a total hydrogen storage of 50 standard cubic meters, an equivalent location of 8000 meters, and an available time window of [600, 600] seconds, then the second cluster would... The final virtual hydrogen reservoir state vector is: .

[0158] S6. Using the virtual hydrogen energy reservoir state vector as the resource point and the hydrogen-using unit corresponding to the load point as the target point, based on the spatiotemporal grid of hydrogen energy transmission, a spatiotemporal matching search is performed in the direction that satisfies the intensity gradient of hydrogen energy demand urgency, and a candidate set of hydrogen energy flow collaborative paths is obtained.

[0159] The specific implementation process of step S6 is as follows:

[0160] S6.1. Using each virtual hydrogen resource cluster in the virtual hydrogen energy reservoir state vector as a resource point and the load point indicated by the urgency of hydrogen energy demand as the target load point, a breadth-first search is performed in the hydrogen energy transmission spatiotemporal grid to obtain a set of alternative transmission paths. Specifically:

[0161] The arrangement process of the candidate set of hydrogen energy flow collaborative paths uses the virtual hydrogen energy pool state vector constructed in step S5 as the set of schedulable resource points. Each resource point corresponds to a virtual hydrogen resource cluster, and its position is determined by the current equivalent set position of the cluster. Characterization. Simultaneously, the load points pointed to in the hydrogen demand urgency list generated in step S4 are used as target points, each target point being associated with its specific demand time and urgency value. The core data foundation is the hydrogen energy transmission spatiotemporal grid generated in step S2, which defines the spatial connectivity between pipeline nodes and the estimated time required for hydrogen to flow through each node. The search guiding principle is to satisfy the direction of the hydrogen demand urgency intensity gradient, that is, to prioritize matching resources to target load points with higher urgency.

[0162] Starting with each virtual hydrogen resource cluster, the search algorithm uses the grid node corresponding to its equivalent settling location as the source node, and performs a breadth-first search within the hydrogen energy transmission spatiotemporal grid. The search algorithm begins at the source node, sequentially visits all its adjacent grid nodes, and calculates the cumulative transmission time from the source node to each visited node based on the estimated arrival time data stored in the grid. This process continues until all grid nodes corresponding to the target load points are visited or the preset maximum search step count is reached, thereby generating a set of alternative transmission paths for each pair of resource points and target points. Each alternative transmission path is recorded as a sequence of grid nodes traversed, resulting in a set of alternative transmission paths. ,in, For the node corresponding to the resource point, The node corresponding to the target load point.

[0163] S6.2. From the set of alternative transmission paths, select the path for which the expected hydrogen concentration at each grid node meets the concentration threshold of the target load point, and accumulate the total transmission delay of each path to obtain the set of feasible paths. Specifically:

[0164] Concentration threshold screening is performed on all candidate transport paths. For each path, every intermediate grid node except the starting point is examined. The projected hydrogen concentration when the resource hydrogen is expected to arrive at this node. This concentration must not be lower than the minimum hydrogen concentration threshold required by the hydrogen unit at the target load point. Will satisfy For all All valid paths are retained, forming a set of feasible paths. For each path in the set of feasible paths, its total transmission delay needs to be accumulated. This delay is the total time it takes for hydrogen to travel from the resource point to the target load point, and its value is the value of the node corresponding to the target load point. Cumulative transmission time The calculation formula is:

[0165] ;

[0166] in, For the spatiotemporal grid of hydrogen energy transfer, from the node To the node neighboring nodes The estimated transmission time. Finally, a demand time window check is performed, and the preset demand time window... The timeout period is set by the dispatching system based on the urgency of the power grid regulation needs, for example, 900 seconds.

[0167] S6.3 Select feasible paths from the set of feasible paths whose total transmission delay is less than the preset demand time window, and bind them with the corresponding virtual hydrogen resource clusters and target load points to form a candidate set of hydrogen energy flow collaborative paths.

[0168] From the set of feasible paths, select those paths with the lowest total transmission delay. The demand time window is less than or equal to the target load point. All paths are identified. These paths, selected through spatiotemporal matching, concentration threshold filtering, and delay checks, are linked to the virtual hydrogen resource cluster that initiated the search and the final target load point, collectively forming a hydrogen energy flow collaborative path record. All such records are then aggregated to form a candidate set of hydrogen energy flow collaborative paths.

[0169] For example, suppose the virtual hydrogen reservoir state vector contains resource point R1, whose equivalent location corresponds to grid node N5. The target load point ranked highest in the hydrogen demand urgency list corresponds to grid node N12, and its demand time window... The minimum hydrogen concentration threshold is 900 seconds. The value is 0.85. A breadth-first search is performed within the hydrogen energy transfer spatiotemporal grid, using N5 as the source node. An alternative path is found. By consulting the grid data, the estimated transmission time for each segment can be obtained: Second, Second, Seconds. Total transmission delay along the path. Seconds. Check concentration: Read grid predictions for the concentration of hydrogen when it reaches N12 in 650 seconds. The concentration at intermediate node N8 is 0.89 after 200 seconds, and the concentration at N10 is 0.875 after 350 seconds. These values ​​are all greater than or equal to... Therefore, screening is performed using a concentration threshold. Because seconds less The path passes the demand time window check in seconds. Therefore, resource point R1, target load point N12, and alternative paths are compared. Binding creates a candidate record. This assumes there is another alternative path for the same pair of resource points and target points. The total transmission delay is 1000 seconds. Although the concentration meets the requirements, if the delay exceeds the required time window of 900 seconds, the path is filtered out and does not enter the candidate set. The system performs the above search and filtering process on all resource points and all high-urgency target points. The final hydrogen energy flow collaborative path candidate set may contain multiple such records, for example: {Resource point: R1, Target point: N12, Path: [N5, N8, N10, N12], Total delay: 650 seconds}; {Resource point: R2, Target point: N15, Path: [N3, N9, N15], Total delay: 520 seconds}.

[0170] S7. With energy cost and power grid balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen doping safety threshold of the gas turbine unit as constraints, establish an objective function, and select the final hydrogen energy flow collaborative path from the candidate set of hydrogen energy flow collaborative paths based on the objective function.

[0171] The specific implementation process of step S7 is as follows:

[0172] S7.1 Establish a cost calculation model that includes hydrogen production costs, pipeline transportation costs, and regulation and compensation costs, as the energy cost target item. Specifically:

[0173] The final screening process for hydrogen energy flow collaborative paths in step S7 is based on the candidate set of hydrogen energy flow collaborative paths arranged in step S6, and two quantitative evaluation targets are established for each candidate path.

[0174] The first objective is energy cost, establishing a cost calculation model that includes hydrogen production costs, pipeline transportation costs, and regulation and compensation costs. Hydrogen production costs are directly proportional to the volume of hydrogen extracted from the source or storage point; pipeline transportation costs are directly proportional to the transportation distance and pipeline resistance; and regulation and compensation costs are an economic incentive or compensation for rapidly adjusting the use of virtual hydrogen resources in transit. Energy cost objective item The specific calculation formula for a candidate path is as follows:

[0175] ;

[0176] in, The production cost per unit volume of hydrogen is expressed in yuan per standard cubic meter, and is based on the operating data of hydrogen production plants and market quotations. The hydrogen volume for the candidate path plan is taken from the total hydrogen inventory of the corresponding virtual hydrogen resource cluster or the portion that meets the demand, in standard cubic meters. The cost coefficient for transporting hydrogen per unit volume per unit pipe length is expressed in yuan per standard cubic meter per kilometer and is based on the gas transmission rate of the pipeline operator. The length of each pipe segment in the path is in kilometers; This is the absolute value of the average flow rate of hydrogen gas in the pipeline segment of the candidate route, in standard cubic meters per hour, used to reflect the impact of gas transmission scale on cost; The compensation cost factor per unit of adjustable power, expressed in yuan per megawatt, is set based on the compensation standards of the grid ancillary services market. The total response time of the candidate path, i.e. the total transmission delay of the path, is expressed in seconds. This item reflects the adjustment speed; the faster the speed, the higher the unit compensation cost may be.

[0177] S7.2. Establish a model that minimizes the sum of real-time power deficit data at each key node of the power grid, as the objective term for power grid power balance. Specifically:

[0178] The second objective is grid power balance, which involves establishing a model to minimize the sum of real-time power deficit data at key nodes in the grid. For a candidate path, its contribution lies in reducing the power deficit associated with that node by supplying hydrogen to hydrogen-consuming units at the target load point. Grid power balance objective item The calculation formula is:

[0179] ;

[0180] in, It is the set of all time points within the scheduling period; The grid node associated with the target load point at time The predicted power deficit is in megawatts. The power-to-hydrogen conversion factor for the hydrogen unit is the power that can be generated or balanced per unit volume of hydrogen, expressed in megawatts per standard cubic meter. Its reciprocal is the power used in step S4. ; For at any time The actual volume of hydrogen reaching the target load point and available for power balancing is the planned dispatch volume. A function based on path transmission delay and hydrogen consumption rate, in standard cubic meters; the smaller the value of this objective term, the greater the contribution of the path to the power balance of the grid, i.e., the greater the reduction of power deficit.

[0181] S7.3. Combining the energy cost objective and the grid power balance objective, calculate the contribution value of the energy cost of the hydrogen-using unit corresponding to each path in the candidate set of hydrogen energy flow cooperative paths to the grid power balance; the contribution value is the solution of multi-objective optimization, i.e., contribution value = ;

[0182] S7.4. The allowable ramp rate of the hydrogen-using unit electrolyzer is used as the time-series variation constraint, and the allowable hydrogen blending ratio of the hydrogen-using unit gas turbine is used as the concentration constraint. Specifically:

[0183] Next, the constraints are addressed. The time-varying constraints require that the rate of change of the input hydrogen flow rate or output power of the hydrogen-using unit at the target load point, especially the electrolyzer, must not exceed the allowable ramp rate of the equipment. ,Right now ,in, This refers to the volumetric flow rate of hydrogen, expressed in standard cubic meters per hour. The unit is standard cubic meters per hour per minute. Concentration constraints require that the hydrogen concentration at the gas turbine unit be within acceptable limits. It must be within its permissible blending ratio range, i.e. ,in, and These are the minimum and maximum hydrogen concentration thresholds required for the hydrogen unit to operate normally at the load point, as predicted by the spatiotemporal grid of hydrogen energy transmission, and are set based on the equipment performance parameters in the equipment safety operation manual.

[0184] S7.5. Based on the time-series variation constraint and the concentration constraint, a multi-objective optimization algorithm is used to balance and solve the contribution value. The Pareto optimal solution is selected from the candidate set of hydrogen energy flow collaborative paths, and determined as the final hydrogen energy flow collaborative path. Specifically:

[0185] A multi-objective optimization algorithm is employed to weigh the contribution values ​​of all paths in the candidate set of hydrogen energy flow collaborative paths. Specifically, each candidate path is mapped to a point in a two-dimensional objective space, with coordinates as follows: The Pareto optimal solution set screening method is employed to identify points from all available points where further improvement is impossible on any one objective without compromising the other. The paths corresponding to these points constitute the Pareto front. Finally, based on the system's preset preferences, such as prioritizing grid balance or controlling costs, a final solution is selected from the Pareto optimal solution set. The path corresponding to this solution is then determined as the final hydrogen energy flow collaborative path.

[0186] For example, suppose the candidate set of hydrogen energy flow collaborative paths contains two paths, Path1 and Path2. Path1 plans to schedule hydrogen volume. Standard cubic meters, originating from virtual resource cluster R1, total transport distance. kilometers, average flow Standard cubic meters per hour, total response time Seconds. Path2 scheduling volume Standard cubic meters, total conveying distance of 30 kilometers, average flow rate of 800 standard cubic meters per hour, total response time of 400 seconds. Set cost coefficient. Yuan per standard cubic meter, Yuan per standard cubic meter per kilometer Yuan per megawatt. Therefore, the energy cost of Path1 is... The figure here is for example calculation; the last item needs to be calculated based on the actual conversion factor to determine the equivalent power. Energy cost of Path2. Yuan. Assuming the target load points of the two paths are the same, the predicted power deficit values ​​of the target load points of the two paths are... With a total of 100 MWh within the scheduling cycle, Path1, due to its larger scheduling volume, is expected to reduce the deficit by 80 MWh, while Path2 is expected to reduce it by 50 MWh. Megawatts per standard cubic meter, then the balance of Path1 Megawatt-hours, Path2 balance Megawatt-hours. Constraint check: Hydrogen concentration in Path1 reaches 0.87, within the allowable blending ratio for the target gas turbine unit. Within the range, the concentration constraint is met; however, its hydrogen flow rate change rate, calculated to be 10 standard cubic meters per hour per minute, does not exceed the electrolyzer's allowable ramp rate of 15 standard cubic meters per hour per minute, thus satisfying the time-series variation constraint. Path2 also satisfies both constraints. In the constructed two-dimensional space, Path1 has a point at (6416.5, 98), and Path2 has a point at (3750, 98.8). Path2 has a lower cost but a slightly weaker contribution to grid balance. If other paths exist, a Pareto comparison is used. Assuming no other path is superior to either Path1 or Path2 in both cost and grid balance, Path1 and Path2 constitute the Pareto front. If the system strategy prioritizes cost, Path2 is selected as the final hydrogen energy flow collaborative path; if grid balance is prioritized, Path1 is selected. Path2 is selected according to preset rules, thus the final hydrogen energy flow collaborative path is determined to be Path2.

[0187] S8. Construct a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decode and generate pipeline scheduling control instructions from the timing control nodes in the scheduling trajectory matrix.

[0188] The specific implementation process of step S8 is as follows:

[0189] S8.1. Deploy the final hydrogen energy flow coordination path according to time series, and generate a scheduling instruction group containing instruction time, action type, and operation quantity for each gas source point, storage point, distribution point, and load point involved in the final hydrogen energy flow coordination path. The action types include pressurization, diversion, injection, and consumption. Specifically:

[0190] Specifically, the continuous transport process in the path is discretized into a sequence of events arranged in chronological order, with each event corresponding to the time point when hydrogen arrives at or flows through a key topological node. A scheduling instruction group is generated for each gas source, storage point, distribution point, and load point involved in the final hydrogen energy flow coordination path. Each scheduling instruction group contains three core fields: instruction time (time stamp based on Coordinated Universal Time, accurate to the second); action type (selected from four preset operations, including pressurization, diversion, injection, and consumption); and operation quantity (a numerical value with units, the specific type and unit determined according to the action type; for example, pressurization corresponds to pressure change in kilopascals, diversion to diversion ratio adjustment, injection to hydrogen volumetric flow rate in standard cubic meters per hour, and consumption to hydrogen consumption rate in standard cubic meters per hour).

[0191] The calculation of the operational quantities is based on the predicted data from the hydrogen energy transfer spatiotemporal grid and the target regulation requirements. For the diversion operation at the distribution point Jn, the diversion ratio adjustment value is... :

[0192] ;

[0193] in, The volumetric flow rate of hydrogen from distribution point Jn to the target pipeline segment according to the final hydrogen energy flow coordination path requirements, expressed in standard cubic meters per hour; This represents the total hydrogen volumetric flow rate upstream of the distribution point Jn; The actual diversion ratio of the distribution point Jn to the target pipeline at the current moment; calculated This refers to the incremental adjustment of the diversion ratio, which is set based on the node flow balance equation. For injection operations at gas source points or storage points, the operational quantity... To meet the flow demand of downstream paths and compensate for pipeline pressure losses, the calculation formula can be simplified as follows:

[0194] ;

[0195] in, The target pressure of downstream nodes required by the path; For the current pressure; The flow-pressure gain coefficient is determined based on the pipeline resistance characteristics, with units of standard cubic meters per hour per kilopascal. Its value is obtained by calibration based on the pipeline hydraulic calculation model.

[0196] S8.2 Arrange all scheduling instruction groups in chronological order and associate them with the topology nodes of the hydrogen-blended integrated energy system pipeline network, constructing a two-dimensional scheduling trajectory matrix. The rows of the scheduling trajectory matrix represent the topology nodes of the hydrogen-blended integrated energy system pipeline network, and the columns represent the time sequence. Specifically:

[0197] After all scheduling instruction groups are generated, they are sorted according to their chronological order, forming a time-indexed instruction list. Simultaneously, these instructions are associated with the topology nodes of the hydrogen-blended integrated energy system pipeline network, with each instruction group corresponding to one topology node. Based on this, a two-dimensional scheduling trajectory matrix is ​​constructed. Row index of the scheduling trajectory matrix Represents the topology node number in the hydrogen-blended integrated energy system pipeline network, column index. Represents the discretized time sequence number, starting from the current time, at fixed time intervals. For example, incrementing every 60 seconds. In the matrix, the row index of a node is its node number; 1, 2, 3 represent node 1, node 2, and node 3 respectively. The column indexes 1, 2, 3 represent time sequence 1, time sequence 2, and time sequence 3 respectively, with a time interval between time sequence 1 and time sequence 2. Matrix elements What is stored is the node In time sequence The corresponding time interval contains the set of scheduling instructions to be executed; if a node has no instructions within a certain time interval, the corresponding element is empty. The process of constructing the matrix is ​​essentially to reorganize the time-series instruction list according to the two dimensions of nodes and time.

[0198] S8.3. Extract the timing control nodes and corresponding control instructions corresponding to the scheduling instruction groups to be executed within the current scheduling period from the scheduling trajectory matrix. Specifically:

[0199] Extract the scheduling instructions that need to be executed immediately within the current scheduling period from the scheduling trajectory matrix. The current scheduling period is defined as starting from the current time. arrive Time window, for example The time window is 300 seconds. Traverse the scheduling trajectory matrix to identify instruction groups whose times fall within this time window, and record the corresponding timing control nodes for these instruction groups. A timing control node is a tuple. ,in, Number the topology nodes. This refers to the timing number to which the control instruction belongs. The extracted control instruction contains complete action type and operation quantity information.

[0200] S8.4. Encode the timing control nodes and control commands into a standard communication protocol format and send them to the corresponding topology nodes of the hydrogen-blended integrated energy system pipeline network to generate pipeline network scheduling control commands.

[0201] Finally, instruction encoding and distribution are performed. The timing control node information and corresponding control instructions are encoded into a standard communication protocol format, such as the IEC61850 MMS protocol or Modbus TCP protocol. The encoding process includes converting node numbers to logical device addresses in the protocol, converting instruction timestamps to protocol timestamps, mapping action types to protocol function codes, and converting operational quantities to values ​​in protocol data units. The encoded instruction data packets are then distributed through the industrial control network to the programmable logic controllers or remote terminal units at the corresponding physical topology nodes in the hydrogen-blended integrated energy system pipeline network. These controllers execute the received instructions to control valves, compressors, injection equipment, or consumable equipment, thereby generating and executing official dispatch control instructions.

[0202] For example, assume the final hydrogen flow coordination path involves a gas source point S1, a distribution point J1, and a load point L1. The path requires hydrogen to arrive at J1 at t=600 seconds and at L1 at t=900 seconds. Calculations show that the diversion ratio at point J1 needs to be adjusted at t=600 seconds to guide the hydrogen flow towards L1. The current diversion ratio at point J1 to the pipeline leading to L1 is... The ratio is 0.3, the path requires this ratio to be 0.5, and the total hydrogen volumetric flow rate upstream of distribution point J1. If the flow rate is 1000 standard cubic meters per hour, then the volumetric flow rate of hydrogen from this distribution point to the target pipeline segment is calculated according to the final hydrogen energy flow coordination path. Standard cubic meters per hour. The required adjustment to the diversion ratio increment. Therefore, a scheduling instruction group is generated for J1: instruction time 600 seconds, action type "diversion", operation quantity +0.2. Simultaneously, to ensure flow rate, a "pressurization" operation needs to be performed on gas source point S1 at t=0 seconds (immediately), with an operation quantity calculated to be +50 kPa. A "consumption" instruction is generated for load point L1 at t=900 seconds, with an operation quantity of 500 standard cubic meters per hour. All instructions are sorted according to their time sequence. Assume the time interval... If the time interval is 300 seconds, then the time is divided into time slots: time slot 1 (0-300 seconds), time slot 2 (300-600 seconds), time slot 3 (600-900 seconds), and time slot 4 (900-1200 seconds). Construct a scheduling trajectory matrix, with each row corresponding to nodes S1, J1, and L1. The matrix elements are: Includes the command "Boost, +50 kPa"; Includes the instruction "Spinning, +0.2"; The system contains the instruction "Consumption, 500 standard cubic meters per hour"; all other elements are empty. The current scheduling cycle is set to 300 seconds, meaning only instructions in time slot 1 are considered. The timing control node (S1,1) is extracted from the matrix, corresponding to the control instruction "Pressure boost, +50 kPa". This instruction is encoded in Modbus TCP protocol format: the target address corresponds to the holding register address of the S1 controller, and the written value is 50 (representing the pressure increment). This instruction is sent through the control network. The compressor at the gas source point S1 receives the instruction and executes the pressure boosting operation, generating the official scheduling control instruction and applying it to the physical pipeline network.

[0203] Example 2

[0204] like Figure 4 As shown, the present invention provides a hydrogen-blended integrated energy system scheduling system that considers dynamic gas flow delay in pipeline networks, comprising:

[0205] The data acquisition module is used to acquire real-time operating data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network, forming a real-time status dataset of the pipeline network;

[0206] The grid generation module is used to generate a spatiotemporal grid for hydrogen energy transmission by calling a pre-trained dynamic simulation model or data-driven model of the pipeline network based on the real-time status dataset of the pipeline network, and performing rolling prediction starting from the current moment.

[0207] The real-time dispatch demand dataset generation module is used to collect power grid frequency deviation data and node power deficit prediction data in real time to obtain the real-time dispatch demand dataset of the power grid.

[0208] The hydrogen energy demand urgency generation module is used to obtain the performance parameters of hydrogen-using units at load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantify the hydrogen energy demand urgency.

[0209] The state vector construction module is used to statistically analyze the hydrogen storage and corresponding gas flow delay information of all hydrogens in the current transmission process in the physical pipeline network based on the connectivity state of the hydrogen transmission spatiotemporal grid, dynamically cluster the discrete hydrogen flow segments, and construct a virtual hydrogen energy pool state vector that can be regarded as a scheduling resource.

[0210] The hydrogen energy flow collaborative path candidate set arrangement module is used to take the virtual hydrogen energy reservoir state vector as the resource point and the hydrogen-using unit corresponding to the load point as the target point, and perform spatiotemporal matching search in the direction that satisfies the hydrogen energy demand urgency intensity gradient according to the hydrogen energy transmission spatiotemporal grid, and arrange it to obtain the hydrogen energy flow collaborative path candidate set.

[0211] The hydrogen energy flow collaborative path determination module is used to establish an objective function with energy cost and grid power balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen doping safety threshold of the gas turbine unit as constraints. Based on the objective function, the module selects the final hydrogen energy flow collaborative path from the candidate set of hydrogen energy flow collaborative paths.

[0212] The control command generation module is used to construct a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decode and generate pipeline scheduling control commands from the timing control nodes in the scheduling trajectory matrix.

[0213] In an optional implementation, the hydrogen-blended integrated energy system scheduling method considering the dynamic airflow delay of the pipeline network includes: a) acquiring real-time operating data of the gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network to form a real-time pipeline network status dataset; b) based on the real-time pipeline network status dataset, calling the pipeline network dynamic simulation model or data-driven model to perform rolling predictions starting from the current moment to generate a hydrogen energy transmission spatiotemporal grid; c) acquiring a real-time grid scheduling demand dataset; d) acquiring the performance parameters of hydrogen-using units, mapping the real-time grid scheduling demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantifying the urgency of hydrogen energy demand; e) constructing a virtual hydrogen energy reservoir state vector that can be regarded as a scheduling resource; f) performing spatiotemporal matching search in the direction that satisfies the intensity gradient of hydrogen energy demand urgency, and arranging to obtain a candidate set of hydrogen energy flow collaborative paths; g) establishing an objective function, and selecting the final hydrogen energy flow collaborative path from the candidate set of hydrogen energy flow collaborative paths based on the objective function; h) constructing a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decoding and generating pipeline network scheduling control instructions from the timing control nodes in the scheduling trajectory matrix.

Claims

1. A method for scheduling a hydrogen-doped integrated energy system considering dynamic gas flow delay in a pipeline network, characterized in that, include: Real-time operational data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network are obtained to form a real-time pipeline network status dataset; Based on the real-time status dataset of the pipeline network, a pre-trained pipeline network dynamic simulation model or data-driven model is invoked to perform rolling predictions starting from the current moment, generating a spatiotemporal grid for hydrogen energy transmission. Real-time acquisition of power grid frequency deviation data and node power deficit prediction data yields a real-time power grid dispatch demand dataset. Obtain the performance parameters of hydrogen-using units at load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantify the urgency of hydrogen energy demand. Based on the connectivity status of the hydrogen energy transmission spatiotemporal grid, the current hydrogen storage and corresponding airflow delay information of all hydrogen in the transmission process within the physical pipeline are statistically analyzed. The discretely distributed hydrogen flow segments are dynamically clustered to construct a virtual hydrogen energy pool state vector that can be regarded as a scheduling resource. Using the virtual hydrogen energy reservoir state vector as resource points and the hydrogen-using units corresponding to the load points as target points, a spatiotemporal matching search is performed in the direction that satisfies the intensity gradient of hydrogen energy demand, based on the spatiotemporal grid of hydrogen energy transmission, and a candidate set of hydrogen energy flow collaborative paths is obtained. With energy cost and grid power balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen blending safety threshold of the gas turbine unit as constraints, an objective function is established, and the final hydrogen energy flow collaborative path is obtained by screening from the candidate set of hydrogen energy flow collaborative paths based on the objective function. Based on the final hydrogen energy flow collaborative path, a scheduling trajectory matrix is ​​constructed, and pipeline scheduling control instructions are generated by decoding the timing control nodes in the scheduling trajectory matrix.

2. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The acquisition of real-time operational data from gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network forms a real-time pipeline network status dataset, including: The hydrogen concentration and flow rate at the gas source point of the hydrogen-blended integrated energy system pipeline network are obtained, and the predicted inlet hydrogen concentration at each downstream node of the pipeline network is calculated by combining the diversion ratio of the hydrogen-blended integrated energy system pipeline network distribution point. Real-time pipeline pressure data and cross-sectional flow data of upstream and downstream adjacent nodes of the hydrogen-blended integrated energy system network are collected simultaneously to calculate the flow-pressure gradient vector characterizing the airflow direction. By combining the predicted inlet hydrogen concentration with the flow-pressure gradient vector, the hydrogen concentration of each node at the next sampling time is calculated, generating node hydrogen concentration parameters that include the spatiotemporal evolution of concentration. The pipeline pressure data, the cross-sectional flow data, and the node hydrogen concentration parameters are aligned in time sequence and associated with the corresponding topological locations in the hydrogen-blended integrated energy system pipeline network, and integrated into a real-time pipeline network status dataset.

3. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The step of generating a hydrogen energy transmission spatiotemporal grid by calling a pre-trained pipeline dynamic simulation model or data-driven model based on the real-time status dataset of the pipeline network and performing rolling predictions starting from the current moment includes: Based on the pipeline pressure data and cross-sectional flow data in the real-time status dataset of the pipeline network, the initial steady-state power flow distribution of the pipeline network at the current moment is calculated using the preset gas pipeline network fluid dynamics equation. Using the node hydrogen concentration parameters in the real-time status dataset of the pipeline network as boundary conditions, the convection and diffusion process of hydrogen under the initial steady-state power flow distribution of the pipeline network is simulated to generate a preliminary hydrogen energy spatiotemporal distribution grid. Based on the cross-sectional flow data, the preliminary hydrogen energy spatiotemporal distribution grid is dynamically and continuously corrected, and the expected hydrogen arrival concentration and expected arrival time of each grid cell in the preliminary hydrogen energy spatiotemporal distribution grid are updated to obtain the hydrogen energy transmission spatiotemporal grid.

4. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The real-time acquisition of power grid frequency deviation data and node power deficit prediction data yields a real-time power grid dispatch demand dataset, including: The actual frequency of each key node in the power grid is monitored in real time, and the deviation between the actual frequency and the preset rated frequency is calculated to obtain power grid frequency deviation data. Acquire renewable energy output forecast data and load forecast data, calculate the power generation and consumption difference of each key node in the power grid within the future dispatch cycle, and generate power deficit forecast data for key nodes. By associating the key nodes whose geographical locations overlap with the power grid frequency deviation data and the key node power deficit prediction data, and performing weighted fusion, a real-time power grid dispatch demand dataset representing the intensity and location of power grid regulation demand is formed.

5. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The process of obtaining the performance parameters of hydrogen-consuming units at load points corresponding to the real-time grid dispatch demand dataset, mapping the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantifying the urgency of hydrogen energy demand includes: Obtain the type of hydrogen-using unit connected to the load point and its power-to-hydrogen-consumption conversion coefficient; Based on the real-time grid dispatch demand dataset, a direct mapping relationship between grid power demand and hydrogen demand is established by combining the power-hydrogen consumption conversion coefficient. The equivalent hydrogen demand required by each load point to balance grid power at different times in the future is calculated through the direct mapping relationship between grid power demand and hydrogen demand. On the spatiotemporal grid of hydrogen energy transmission, the grid nodes corresponding to each load point are located, and the hydrogen supply gap intensity under time constraints is calculated based on the equivalent hydrogen demand and the expected hydrogen arrival concentration and expected arrival time of the grid nodes corresponding to each load point. The intensity of the hydrogen supply gap at all load points is normalized and sorted to generate a hydrogen demand urgency rating that characterizes the urgency and magnitude of demand at each grid node.

6. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The process involves statistically analyzing the hydrogen storage capacity and corresponding gas flow delay information of all hydrogen gas undergoing transmission within the physical pipeline network at the current moment, based on the connectivity state of the hydrogen energy transmission spatiotemporal grid. This is followed by dynamically clustering the discretely distributed hydrogen flow segments to construct a virtual hydrogen reservoir state vector that can be considered as a scheduling resource. Analyze the spatiotemporal grid of hydrogen energy transmission to identify all hydrogen flow segments that have been sent from the gas source or storage point at the current moment but have not yet arrived at the target load point; Extract the current spatial location, remaining transport distance, current airflow speed, and hydrogen inventory of each hydrogen flow segment in transit, calculate the estimated time to reach the target load point, and obtain airflow delay information; Based on the gas flow delay information and the adjacency relationship of the flow segments in the pipeline network topology of the hydrogen-infused integrated energy system, multiple discrete hydrogen flow segments with similar expected arrival times and located in continuous pipeline segments are aggregated to form multiple virtual hydrogen resource clusters. Each virtual hydrogen resource cluster is assigned attribute information, and all virtual hydrogen resource clusters together constitute a virtual hydrogen energy pool state vector. The attribute information includes the total hydrogen storage within the cluster, the current equivalent concentration location, and the available time window.

7. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The virtual hydrogen reservoir state vector is used as a resource point, and the hydrogen-using unit corresponding to the load point is used as the target point. Based on the hydrogen energy transmission spatiotemporal grid, a spatiotemporal matching search is performed in the direction that satisfies the intensity gradient of hydrogen energy demand urgency, and a candidate set of hydrogen energy flow cooperative paths is obtained, including: Using each virtual hydrogen resource cluster in the virtual hydrogen energy reservoir state vector as a resource point and the load point pointed to by the hydrogen energy demand urgency as the target load point, a breadth-first search is performed in the hydrogen energy transmission spatiotemporal grid to obtain a set of alternative transmission paths. From the set of alternative transmission paths, select the path whose expected hydrogen concentration at each grid node meets the concentration threshold of the target load point, and accumulate the total transmission delay of the path to obtain the set of feasible paths. From the set of feasible paths, feasible paths with a total transmission delay less than the preset demand time window are selected and bound to the corresponding virtual hydrogen resource clusters and target load points to form a candidate set of hydrogen energy flow collaborative paths.

8. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The objective function is established with energy cost and grid power balance as targets, and the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen blending safety threshold of the gas turbine unit as constraints. Based on this objective function, the final hydrogen energy flow collaborative path is obtained by screening from the candidate set of hydrogen energy flow collaborative paths, including: Establish a cost calculation model that includes hydrogen production costs, pipeline transportation costs, and regulation and compensation costs, as the target item for energy costs; A model is established to minimize the sum of real-time power deficit data at each key node of the power grid, which serves as the target term for the power grid power balance. Combining the energy cost target and the power grid balance target, the contribution value of the energy cost and the power grid balance of the hydrogen-using unit corresponding to each path in the candidate set of hydrogen energy flow collaborative paths is calculated. The allowable ramp rate of the hydrogen-using unit electrolyzer is used as the time-series variation constraint, and the allowable hydrogen blending ratio of the hydrogen-using unit gas turbine is used as the concentration constraint. Based on the time-series variation constraints and the concentration constraints, a multi-objective optimization algorithm is used to weigh and solve the contribution value, and the Pareto optimal solution is selected from the candidate set of hydrogen energy flow collaborative paths to determine the final hydrogen energy flow collaborative path.

9. The hydrogen-doped integrated energy system scheduling method considering dynamic gas flow delay in the pipeline network according to claim 1, characterized in that, The process of constructing a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decoding and generating pipeline scheduling control instructions from the timing control nodes in the scheduling trajectory matrix, includes: The final hydrogen energy flow coordination path is unfolded according to the time sequence, and a scheduling instruction group containing instruction time, action type and operation amount is generated for each gas source point, storage point, distribution point and load point involved in the final hydrogen energy flow coordination path. The action type includes pressurization, diversion, injection and consumption. All scheduling instruction groups are arranged in chronological order and associated with the topology nodes of the hydrogen-blended integrated energy system pipeline network to construct a two-dimensional scheduling trajectory matrix. The rows of the scheduling trajectory matrix represent the topology nodes of the hydrogen-blended integrated energy system pipeline network, and the columns represent the time sequence. Extract the timing control nodes and corresponding control instructions corresponding to the scheduling instruction groups to be executed in the current scheduling period from the scheduling trajectory matrix; The timing control node and the control command are encoded into a standard communication protocol format and sent to the topology node corresponding to the hydrogen-blended integrated energy system pipeline network to generate pipeline network scheduling control commands.

10. A hydrogen-doped integrated energy system scheduling system considering dynamic gas flow delay in pipeline networks, characterized in that, include: The data acquisition module is used to acquire real-time operating data of gas source points, storage points, distribution points, and load points of the hydrogen-blended integrated energy system pipeline network, forming a real-time status dataset of the pipeline network; The grid generation module is used to generate a spatiotemporal grid for hydrogen energy transmission by calling a pre-trained dynamic simulation model or data-driven model of the pipeline network based on the real-time status dataset of the pipeline network, and performing rolling prediction starting from the current moment. The real-time dispatch demand dataset generation module is used to collect power grid frequency deviation data and node power deficit prediction data in real time to obtain the real-time dispatch demand dataset of the power grid. The hydrogen energy demand urgency generation module is used to obtain the performance parameters of hydrogen-using units at load points corresponding to the real-time grid dispatch demand dataset, map the real-time grid dispatch demand dataset to the hydrogen energy transmission spatiotemporal grid, and quantify the hydrogen energy demand urgency. The state vector construction module is used to statistically analyze the hydrogen storage and corresponding gas flow delay information of all hydrogens in the current transmission process in the physical pipeline network based on the connectivity state of the hydrogen transmission spatiotemporal grid, dynamically cluster the discrete hydrogen flow segments, and construct a virtual hydrogen energy pool state vector that can be regarded as a scheduling resource. The hydrogen energy flow collaborative path candidate set arrangement module is used to take the virtual hydrogen energy reservoir state vector as the resource point and the hydrogen-using unit corresponding to the load point as the target point, and perform spatiotemporal matching search in the direction that satisfies the hydrogen energy demand urgency intensity gradient according to the hydrogen energy transmission spatiotemporal grid, and arrange it to obtain the hydrogen energy flow collaborative path candidate set. The hydrogen energy flow collaborative path determination module is used to establish an objective function with energy cost and grid power balance as objectives, and with the electrolyzer response rate of the hydrogen-using unit at the corresponding load point and the hydrogen doping safety threshold of the gas turbine unit as constraints. Based on the objective function, the module selects the final hydrogen energy flow collaborative path from the candidate set of hydrogen energy flow collaborative paths. The control command generation module is used to construct a scheduling trajectory matrix based on the final hydrogen energy flow collaborative path, and decode and generate pipeline scheduling control commands from the timing control nodes in the scheduling trajectory matrix.

Images