Multi-energy network scheduling method based on centralized and distributed double-layer optimization strategy

Abstract

The application discloses a kind of multi-energy network scheduling methods based on centralized and distributed double-layer optimization strategy, comprising: collecting port weather data, according to port operation task load historical data to obtain prediction data by prediction;According to prediction data, the multi-commodity flow model of multi-energy network is established;The distributed energy output data in the whole energy distribution period is calculated;The optimal centralized energy distribution scheme is selected and energy distribution is carried out;Optimize starting node and optimize node sequence, determine the error value of real-time data and prediction data, determine the actual output and load data of each type of energy according to error value, and update starting node data;Optimize target node and adjacent node energy distribution scheme to obtain optimization result, and replace centralized energy distribution scheme with optimization result;Adjust energy distribution.The application guarantees the feasibility, reliability and robustness of energy distribution scheme, and makes full use of energy facilities and computing resources in energy network.

Description

Technical Field

[0001] This invention belongs to the field of port energy dispatching, specifically involving a multi-energy network dispatching method based on a centralized and distributed dual-layer optimization strategy. Background Technology

[0002] With the continuous expansion of port operations and the increasing automation and intelligence of ports, port production processes and the energy structure of related equipment are becoming increasingly complex. The crucial role of energy dispatch in improving port operational efficiency and reducing energy consumption is becoming increasingly prominent. Effective optimization of port energy consumption requires scientific planning of the port energy network system and effective collaboration with technologies such as artificial intelligence and operations optimization to improve the coordinated allocation and emergency response capabilities of the port energy network. Compared with traditional centralized power grid structures, distributed energy has advantages such as being clean and environmentally friendly, having flexible power generation methods, and low maintenance costs. However, the large-scale grid connection of distributed energy sources also poses significant challenges to the stability and reliability of the power grid.

[0003] As an autonomous power system, smart microgrids can efficiently match supply and demand within distributed energy networks. Smart trucks, quay cranes, and yard cranes, as the main energy-consuming loads in port operations, are also special distributed energy sources, possessing the dual attributes of controllable loads and power sources, and exhibiting characteristics such as mobility and randomness. Based on port operation equipment, the flexibility and distribution of smart microgrids, along with the integrated characteristics of charging, discharging, and storage power stations, ensure the stability, availability, and responsiveness of the distributed energy network, laying the foundation for further optimization of port distributed energy scheduling and control.

[0004] Currently, energy networks generally adopt a centralized structure and energy dispatch technology. These traditional dispatch methods require global information collected by a centralized computing center under the centralized energy network layout. Their low flexibility, low scalability, high communication costs, and high computing costs make it difficult for these methods to be directly applied in distributed energy networks that have been connected to a large number of smart microgrids. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy. This method can comprehensively consider every task in the port production process, call all allocated energy objects to participate in the energy allocation scheme, achieve a good balance between the overall optimization and local optimization of the energy allocation scheme, and ensure the feasibility, reliability and robustness of the energy allocation scheme through data prediction technology and consistency verification technology, making full use of the energy facilities and computing resources within the energy network.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is: a multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy, comprising the following steps:

[0007] S1. Collect port weather data including port sunshine conditions and wind speed conditions, and make predictions based on historical data of port operation task load. The prediction data includes the total load data of the energy network in the next energy distribution cycle and the output data of various energy sources such as port wind power and photovoltaic power.

[0008] S2. Based on the predicted data, and with the goal of maximizing energy revenue, a multi-commodity flow model of a multi-energy network is established under the constraint conditions to reflect the coupling between the port logistics network and the energy network. Through operations research optimization techniques, the distributed energy output data for each operational task within the container port is calculated to obtain the data for the entire energy allocation cycle. A container port traffic-energy network coupling simulation model is established, and the optimal centralized energy allocation scheme is selected based on the simulation verification results. Before the real-time energy allocation time arrives, each energy microgrid allocates energy according to the optimal centralized energy allocation scheme. The constraints include flow balance constraints, power constraints, capacity constraints, and energy conservation constraints.

[0009] S3. When a certain real-time energy allocation moment is reached, the distributed energy microgrid detects real-time output and load data; based on the energy network communication topology and port operation task process relationship, it determines the optimized starting node and optimized node sequence of the distributed system, determines the error value between the real-time data and the predicted data at that moment, and determines the actual output and load data of various energy sources in the distributed energy microgrid based on the error value, and updates the starting node data.

[0010] S4. Starting from the initial node of the distributed energy microgrid, according to the adjacency matrix determined by the energy network communication topology and port operation task process relationship, the distributed microgrid nodes in the energy network are traversed sequentially in a certain order. The distributed energy allocation scheme within each node is calculated, and the energy allocation scheme between the target node and its adjacent nodes is optimized to obtain the optimization result. The optimization result is then used to replace the centralized energy allocation scheme. After traversing all nodes in the distributed energy network, a convergence check is performed. If convergence is not achieved, S4 is repeated; otherwise, proceed to S5.

[0011] S5. Each energy microgrid node adjusts energy allocation according to the distributed energy allocation scheme; determines whether the energy allocation cycle has ended. If it has not ended, it waits for the next real-time allocation time and jumps to S3; otherwise, it ends.

[0012] The specific process for obtaining distributed energy output data throughout the entire energy allocation cycle in S2 is as follows:

[0013] S21. Initialize the network, making all arc flows in the multi-commodity flow model zero, and form a cost network according to the arc energy cost.

[0014] S22. Find the shortest path from the source to the receiver for all types of energy on the cost network;

[0015] S23. Adjust the arc flow rate on the shortest path according to the energy demand of the equipment;

[0016] S24. Determine whether the arc flow rate meets the flow balance constraint condition. If it does, proceed to S25; otherwise, jump to S21.

[0017] S25. Calculate the network energy cost at this time;

[0018] S26. Find the longest and shortest paths on the cost network. Calculate whether the difference between the longest and shortest paths has reached the convergence precision. If it has, the optimal flow has been obtained and the current result is output; otherwise, jump to S25.

[0019] The container port traffic-energy network coupled simulation model includes the following characteristics:

[0020] (1) Different types of energy are transmitted through the integrated energy bus. The balance between the energy demand of the equipment and the input energy flow can be represented as the balance of different energy vectors.

[0021] (2) All energy supply, storage and consumption equipment can access the corresponding Sum module by selecting different input energy integrated bus signals, and then perform calculations on the input and output of different energy flows;

[0022] (3) Each device port simulation model has different energy branches according to different energy supply, storage and consumption structures;

[0023] (4) The simulation model realizes the coupling of the traffic network and the energy network through the equipment module and the energy module. The equipment module divides the energy bus signal into different types of energy branches through the Bus Selector module. The energy module realizes the calculation of energy input and output by connecting various types of energy flow to the corresponding Sum module.

[0024] (5) The simulation model calculates the overall load of various energy flow types through the Bus Creator module and displays the output values ​​through the Display module.

[0025] The objective function for the optimal centralized energy distribution scheme is:

[0026] F = max(F3 - F1 - F2)

[0027]

[0028]

[0029]

[0030] In the formula, F represents the network energy cost; F1 represents the electricity cost of the terminal equipment, which consists of the electricity cost of the State Grid, the electricity cost of renewable energy, and the discharge cost of distributed energy storage equipment; F2 represents the terminal's curtailment cost, which consists of the curtailment cost of green energy; F3 represents the terminal's green energy profitability, which consists of the profitability cost of green energy; N represents the number of time periods divided in the scheduling cycle; t represents the current scheduling time period; f represents the energy generation cost function; P represents the energy generation output; f NET f WT f PV f H f LNG f fc , These represent the energy costs of power generation, wind power, solar power, hydrogen energy, liquefied natural gas, centralized energy storage equipment, and distributed energy storage equipment, respectively; P NET P H P LNG , These represent the power generation outputs of the State Grid, hydrogen energy, liquefied natural gas, energy storage equipment, and distributed energy storage equipment, respectively; P WT1 P PV1 P WT2 P PV2 P WT3 P PV3 These represent port wind power consumption, port photovoltaic power consumption, port wind curtailment power consumption, port curtailment solar power consumption, wind power connected to the national grid, and photovoltaic power connected to the national grid, respectively. Among them, NET, WT, PV, H, LNG, and fc represent gray electricity connected to the national grid, distributed wind power, distributed photovoltaic power, hydrogen energy, liquefied natural gas, and energy storage, respectively. k represents distributed energy storage equipment, DIS represents distributed energy storage equipment, out represents discharging, and in represents charging.

[0031] The specific constraints on centralized energy distribution in S2 are as follows:

[0032] (1) Flow balance constraint

[0033]

[0034] P PV (t)=∑ i (P PV1，i (t)+P PV2，i (t)+P PV3，i (t))

[0035] P WT(t)=∑ i (P WT1,i (t)+P WT2,i (t)+P WT3，i (t))

[0036] Flow balance constraints consider the flow balance between supply and demand at node i in the energy network, where load represents the load in the port energy network; where P NET，i P PV1,i P WT1，i P H，i P LNG,i , P represents the power output of node i from power generation, photovoltaic, wind power, hydrogen energy, liquefied natural gas, distributed energy storage devices, and energy storage devices; load，i , P represents the load of node i, the charging of distributed energy storage devices, and the charging of energy storage devices; PV P WT This indicates the total power generation output of photovoltaic and wind power;

[0037] (2) Power Constraint

[0038]

[0039]

[0040] 0≤P PV (t)≤P PV,max

[0041] 0≤P WT (t)≤P WT，max

[0042] 0≤P LNG (t)≤P LNG，max

[0043] 0≤P H (t)≤P H，max

[0044]

[0045]

[0046] in P PV，max P WT，max P LNG,max P H，max P fc，max These represent the rated charging and discharging power of distributed energy storage devices, distributed photovoltaic power, distributed wind power, liquefied natural gas devices, hydrogen energy devices, and energy storage devices, respectively.

[0047] (3) Capacity constraints

[0048]

[0049] in These represent the lower and upper limits of the energy storage system capacity, respectively.

[0050] (4) Energy conservation constraint

[0051]

[0052] This formula indicates that the energy consumption of a distributed energy storage device is less than or equal to the energy storage.

[0053] The formula for calculating the error value in S3 is:

[0054]

[0055] In the formula, ΔP t Q represents the error value within time period t; t This represents the number of nodes in the topology graph within time period t; This represents the predicted state of node i within time period t, where 1 indicates a used state and 0 indicates an unused state. and These represent the actual value and the predicted value of node i within time period t, respectively.

[0056] The formula for calculating the starting node data is:

[0057]

[0058] In the formula, P t，leader The data is the starting node data and is continuously updated as each node is optimized during the optimization process; Predict data for the starting node.

[0059] The specific optimization method for optimizing the energy allocation scheme between the target node and its neighboring nodes in S4 is as follows:

[0060]

[0061] In the formula, This represents the energy consumption revenue of node i during time period t. It is an adjacency matrix. This indicates that there is a connection between node i and node j. This indicates that there is no connection between node i and node j, or i = j.

[0062] The distribution of distributed energy resources in S4 also follows constraints, which are as follows:

[0063] (1) Flow balance constraint

[0064]

[0065] (2) Power Constraint

[0066]

[0067]

[0068] 0≤P PV (t)≤P PV，max

[0069] 0≤P WT (t)≤P WT，max

[0070] 0≤P LNG (t)≤P LNG，max

[0071] 0≤P H (t)≤P H，max

[0072]

[0073]

[0074] (3) Capacity constraints

[0075]

[0076] (4) Energy conservation constraint

[0077]

[0078] After each node is calculated once, the original... and its adjacent nodes The replacement is the corresponding optimization result

[0079] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0080] 1. The centralized layer of this invention pre-allocates energy to the port energy microgrid based on predicted data of port operation tasks, total load of the port energy network, and power generation output of various energy sources within the port. Based on reliable predicted data, it obtains energy allocation results that meet the constraints, and verifies and selects the optimal energy allocation scheme through simulation. Its preliminary scheduling scheme has reliability.

[0081] 2. The distributed optimization and energy dispatching method proposed in this invention can make full use of the computing units and sensing devices under the distributed energy network, and can respond more quickly to changes in energy demand in the microgrid. It has the characteristics of easy implementation, low maintenance cost, high energy efficiency and high robustness.

[0082] 3. The distributed energy network structure proposed in this invention is based on port operation equipment and tasks. By leveraging the flexibility and distribution of smart microgrids and the integrated characteristics of charging, discharging and storage power stations, the coupling between the energy network and the logistics network is strengthened, ensuring the stability, availability and responsiveness of the distributed energy network.

[0083] 4. The distributed scheduling scheme optimization process proposed in this invention can respond promptly to changes in energy consumption during port operations without readjusting the energy allocation scheme of the entire energy network, effectively saving computing power and reducing the computational burden on the energy scheduling system. Attached Figure Description

[0084] Figure 1 This is a flowchart of an embodiment of the present invention;

[0085] Figure 2 This is a simplified multi-commodity flow model conceptual diagram of a port multi-energy network in an embodiment of the present invention;

[0086] Figure 3 This is a flowchart of a port multi-energy network scheduling method based on a distributed strategy in an embodiment of the present invention;

[0087] Figure 4 This is a framework diagram of the container port traffic-energy coupling simulation model in an embodiment of the present invention;

[0088] Figure 5 This is a schematic diagram of the distributed energy network structure of an automated container port in an embodiment of the present invention. Detailed Implementation

[0089] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0090] This invention proposes a multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy. This method comprehensively considers every energy-consuming device in an automated container port, calls on all energy objects to participate in the energy allocation scheme, and achieves a good balance between the overall optimization and local optimization of the energy allocation scheme. Furthermore, it ensures the feasibility, reliability, and robustness of the energy allocation scheme through data prediction technology, consistency verification technology, and simulation verification technology, making full use of the energy facilities and computing resources within the multi-energy network.

[0091] The method divides the energy allocation process into a centralized energy allocation layer based on predictive data and a distributed energy allocation layer based on real-time data, specifically including the following steps:

[0092] S1: The multi-energy network centralized control center collects historical data on port operation load and port weather data (including data related to renewable energy power generation such as wind speed and solar radiation intensity). Based on the historical data prediction information, it uses data prediction methods such as reinforcement learning to predict the total load data of the energy network and the output data of various energy sources in the next energy allocation cycle.

[0093] S2: The centralized control center of the multi-energy network, based on predicted data, establishes a multi-commodity flow model of the multi-energy network with the goal of maximizing energy consumption revenue, while considering constraints such as energy balance and power constraints. For example... Figure 2 As shown, this model refers to the process of abstracting port source, storage, and load facilities into network nodes based on the layout of port production equipment, the energy consumption structure of equipment, and the port energy network. The connection relationship between nodes reflected by the coupling relationship of facilities and the energy consumption structure is abstracted into network edges. The energy consumption demand of equipment is transformed into network edge flow, and the energy consumption revenue is used as the cost of commodity flow, thereby establishing a corresponding multi-commodity flow model.

[0094] Using operations research optimization techniques such as network flow and original duality, the distributed energy output over the entire energy allocation cycle is obtained for each operational task within a container port, such as... Figure 3 As shown, the solution steps are as follows:

[0095] S21. Initialize the network. Set all arc flows in the network model to zero, and form a cost network based on arc energy costs;

[0096] S22. Find the shortest path. Find the shortest path from the source to the sink for each type of energy source on the cost network;

[0097] S23. Adjust the arc flow rate. Adjust the arc flow rate on the shortest path according to the equipment's energy demand;

[0098] S24. Feasibility Determination. Determine whether the arc flow rate satisfies the flow balance constraint. If it does, proceed to the next step; otherwise, repeat S21.

[0099] S25. Calculate energy costs. Calculate the network energy cost F at this time;

[0100] S26. Convergence Check. Find the longest and shortest paths on the cost network, and calculate whether the difference between the longest and shortest paths reaches the convergence precision. If it does, the optimal flow has been found, and the current result is output; otherwise, repeat S25.

[0101] Subsequently, a coupled simulation model of the container port traffic-energy network was established, and the optimal centralized energy distribution scheme was selected based on the simulation verification results. This simulation model was constructed using Simulink, with both equipment and energy modules represented as subsystems. Simulation functional modules were implemented using functional modules from the module library, such as... Figure 4 As shown. The main features of this simulation model are as follows:

[0102] (1) Different types of energy are transmitted through the integrated energy bus. The balance between the energy demand of the equipment and the input energy flow can be represented as the balance of different energy vectors.

[0103] (2) All energy supply, storage and consumption equipment can access the corresponding Sum module by selecting different input energy integrated bus signals, and then perform calculations on the input and output of different energy flows;

[0104] (3) Each device port simulation model has different energy branches according to different energy supply, storage and consumption structures;

[0105] (4) The simulation model realizes the coupling of the traffic network and the energy network through the equipment module and the energy module. The equipment module divides the energy bus signal into different types of energy branches through the Bus Selector module. The energy module realizes the calculation of energy input and output by connecting various types of energy flow to the corresponding Sum module.

[0106] (5) The simulation model calculates the overall load of various energy flow types through the Bus Creator module and displays the output values ​​through the Display module.

[0107] Before the real-time energy allocation occurs, each microgrid distributes energy according to this centralized energy allocation scheme. Its objective function is:

[0108] F = max(F3 - F1 - F2)

[0109]

[0110]

[0111]

[0112] Where F1 represents the electricity cost of terminal equipment, which consists of the electricity cost of the State Grid, the electricity cost of renewable energy, and the discharge cost of distributed energy storage equipment; F2 represents the cost of abandoned electricity at the terminal, which consists of the cost of abandoned electricity from green energy; F3 represents the profitability of green energy at the terminal, which consists of the profit cost of green energy; N represents the number of time periods divided in the scheduling cycle; f represents the energy generation cost function; P represents the energy generation output; NET, WT, PV, H, LNG, and fc represent gray electricity, distributed wind power, distributed photovoltaic power, hydrogen energy, liquefied natural gas, and energy storage connected to the State Grid, respectively; k represents distributed energy storage equipment; DIS represents distributed energy storage equipment; out represents discharge; in represents charging; PV1 represents photovoltaic electricity consumption at the port; WT1 represents wind power electricity consumption at the port; PV2 represents abandoned photovoltaic electricity consumption at the port; WT2 represents abandoned wind power electricity consumption at the port; PV3 represents photovoltaic electricity connected to the State Grid; and WT3 represents wind power connected to the State Grid.

[0113] The constraints governing centralized energy distribution are as follows:

[0114] (1) Flow balance constraint

[0115]

[0116] P PV (t)=∑ i (P PV1,i (t)+P PV2,i (t)+P PV3,i (t))

[0117] P WT (t)=∑ i (P WT1，i (t)+P WT2，i (t)+P WT3，i (t))

[0118] The flow balance constraint considers the flow balance between supply and demand at node i in the energy network, where load represents the load in the port energy network.

[0119] (2) Power Constraint

[0120]

[0121]

[0122] 0≤P PV (t)≤P PV，max

[0123] 0≤P WT (t)≤P WT，max

[0124] 0≤P LNG(t)≤P LNG，max

[0125] 0≤P H (t)≤P H，max

[0126]

[0127]

[0128] in P PV，max P WT，max P LNG，max P H，max P fc，max These represent the rated power for charging and discharging of distributed energy storage equipment, distributed photovoltaic power, distributed wind power, LNG (liquefied natural gas) equipment, hydrogen energy equipment, and energy storage equipment, respectively.

[0129] (3) Capacity constraints

[0130]

[0131] in These represent the lower and upper limits of the energy storage system capacity, respectively.

[0132] (4) Energy conservation constraint

[0133]

[0134] The energy conservation constraint means that the energy consumption of a distributed energy storage device is less than the energy storage.

[0135] S3: When a certain real-time energy allocation moment is reached, the distributed energy microgrid detects real-time output and load data. Based on the specific network communication topology and port operation task sequence, the starting node and optimization node sequence for distributed optimization are determined. The error value between the real-time data and the predicted data at that moment is determined, and the predicted value of the starting node is corrected based on the error value. The formula for calculating the error value is as follows:

[0136]

[0137] Among them, the error value ΔP t Q represents the error value within time period t; t This represents the number of nodes in the topology graph within time period t; μ represents the predicted state of node i in time period t, where 1 indicates a used state and 0 indicates an unused state; and These represent the actual value and the predicted value of node i within time period t, respectively.

[0138] The actual output and load data of various energy sources in the distributed energy microgrid are determined based on the above formula, and the starting node data is updated. The calculation formula is as follows:

[0139]

[0140] Among them, P t，leader It can be viewed as a global variable, which is continuously updated as each node is optimized during the optimization process.

[0141] S4: Starting from the initial node of the distributed energy microgrid, according to the adjacency matrix determined by the energy network communication topology and port operation task relationships, the distributed microgrid nodes within the energy network are traversed sequentially according to a certain order. Distributed energy allocation schemes within each node are calculated, optimizing the energy allocation scheme between the target node and its adjacent nodes. The optimized result replaces the centralized energy allocation scheme. The objective function of this optimization method is:

[0142]

[0143] in, μ represents the energy consumption revenue of node i during time period t. It is an adjacency matrix. This indicates that there is a connection between node i and node j; a value of 0 indicates no connection, or i = j.

[0144] The constraints followed in the distributed energy allocation process are as follows:

[0145] (1) Flow balance constraint

[0146]

[0147] (2) Power Constraint

[0148]

[0149]

[0150] 0≤P PV (t)≤P PV，max

[0151] 0≤P WT (t)≤P WT，max

[0152] 0≤P LNG (t)≤P LNG，max

[0153] 0≤P H (t)≤P H，max

[0154]

[0155]

[0156] (3) Capacity constraints

[0157]

[0158] (4) Energy conservation constraint

[0159]

[0160] After each node is calculated once, its optimized result replaces the original one. and its adjacent nodes After traversing all nodes in the distributed energy network, a convergence check is performed. If convergence is not achieved, step 4 is repeated; otherwise, proceed to the next step.

[0161] S5: Each energy microgrid node adjusts energy allocation according to the distributed energy allocation scheme. It determines whether the energy allocation cycle has ended. If not, it waits for the next real-time allocation time and proceeds to step 3; otherwise, it ends.

[0162] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy, characterized in that, Includes the following steps: S1. Collect port weather data and make predictions based on historical port operation load data. The prediction data includes the total load data of the energy network and the output data of various energy sources within the next energy allocation cycle. The port weather data includes at least the port sunshine conditions and wind speed conditions. The various energy sources include at least port wind power and photovoltaic power. S2. Based on the predicted data, with the goal of maximizing energy revenue, a multi-commodity flow model of a multi-energy network is established under the condition of satisfying the constraints to reflect the coupling between the port logistics network and the energy network; through operations research optimization technology, the distributed energy output data of each operation task in the container port is calculated to obtain the distributed energy output data in the entire energy allocation cycle. Establish a coupled simulation model of container port traffic-energy network, and select the optimal centralized energy distribution scheme based on the simulation verification results; Before the real-time energy allocation time arrives, each energy microgrid allocates energy according to the optimal centralized energy allocation scheme; the constraints include flow balance constraints, power constraints, capacity constraints and energy conservation constraints. S3. When a certain real-time energy allocation moment is reached, the distributed energy microgrid detects real-time output and load data; based on the energy network communication topology and port operation task process relationship, it determines the optimized starting node and optimized node sequence of the distributed system, determines the error value between the real-time data and the predicted data at that moment, and determines the actual output and load data of various energy sources in the distributed energy microgrid based on the error value, and updates the starting node data. S4. Starting from the starting node of the distributed energy microgrid, according to the adjacency matrix determined by the energy network communication topology and port operation task process relationship, the distributed microgrid nodes in the energy network are traversed in a certain order, and the distributed energy allocation scheme within the node is calculated. The energy allocation scheme between the target node and adjacent nodes is optimized to obtain the optimization result, and the optimization result is used to replace the centralized energy allocation scheme. After traversing all nodes in the distributed energy network, a convergence check is performed. If convergence is not achieved, S4 is repeated; otherwise, proceed to S5. S5. Each energy microgrid node adjusts energy allocation according to the distributed energy allocation scheme; determines whether the energy allocation cycle has ended. If it has not ended, it waits for the next real-time allocation time and jumps to S3; otherwise, it ends.

2. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 1, characterized in that, The specific process for obtaining distributed energy output data throughout the entire energy allocation cycle in S2 is as follows: S21. Initialize the network, making all arc flows in the multi-commodity flow model zero, and form a cost network according to the arc energy cost. S22. Find the shortest path from the source to the receiver for all types of energy on the cost network; S23. Adjust the arc flow rate on the shortest path according to the energy demand of the equipment; S24. Determine whether the arc flow rate meets the flow balance constraint condition. If it does, proceed to S25; otherwise, jump to S21. S25. Calculate the network energy cost at this time; S26. Find the longest and shortest paths on the cost network. Calculate whether the difference between the longest and shortest paths has reached the convergence precision. If it has, the optimal flow has been obtained and the current result is output; otherwise, jump to S25.

3. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 2, characterized in that, The container port traffic-energy network coupled simulation model includes the following characteristics: (1) Different types of energy are transmitted through the integrated energy bus. The balance between the energy demand of the equipment and the input energy flow can be represented as the balance of different energy vectors. (2) All energy supply, storage and consumption equipment can access the corresponding Sum module by selecting different input energy integrated bus signals, and then perform calculations on the input and output of different energy flows; (3) Each device port simulation model has different energy branches according to different energy supply, storage and consumption structures; (4) The simulation model realizes the coupling of the traffic network and the energy network through the equipment module and the energy module. The equipment module divides the energy bus signal into different types of energy branches through the Bus Selector module. The energy module realizes the calculation of energy input and output by connecting various types of energy flow to the corresponding Sum module. (5) The simulation model calculates the total load of various energy flow types through the Bus Creator module and displays the output values ​​through the Display module.

4. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 3, characterized in that, The objective function for the optimal centralized energy distribution scheme is: In the formula, The electricity cost of the port equipment consists of the electricity cost from the State Grid, the electricity cost from renewable energy sources, and the discharge cost of distributed energy storage equipment. This indicates the cost of abandoning electricity at the port, which consists of the cost of abandoning green energy. The profitability of green energy at the port is comprised of the cost of generating profits from green energy. N This indicates the number of time periods divided into the scheduling cycle. Indicates the current scheduling time period. f The power generation cost function represents the energy source. P Indicates the power generation output of energy. , , , , , , These represent the energy costs of power generation, wind power, photovoltaic power, hydrogen energy, liquefied natural gas, centralized energy storage equipment, and distributed energy storage equipment, respectively. , , , , These represent the power generation output of China's state-owned power grid, hydrogen energy, liquefied natural gas, energy storage equipment, and distributed energy storage equipment, respectively. , , , , These respectively represent port wind power consumption, port photovoltaic power consumption, port wind curtailment power, port solar curtailment power, wind power connected to the national grid, and photovoltaic power connected to the national grid; among which... NET , WT , PV , H , LNG, FC These respectively represent gray electricity connected to the State Grid, distributed wind power, distributed photovoltaic power, hydrogen energy, liquefied natural gas, and energy storage. k Indicates distributed energy storage devices. DIS This indicates a distributed energy storage device; "out" indicates discharging, and "in" indicates charging.

5. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 4, characterized in that, The specific constraints on centralized energy distribution in S2 are as follows: (1) Flow balance constraint Flow balance constraints consider nodes in the energy network i The supply and demand flow balance, where load represents the load in the port's energy network; among which... , , , , , , Represents a node Power generation output from power plants, photovoltaic, wind power, hydrogen energy, liquefied natural gas, distributed energy storage equipment, and energy storage equipment; , , Represents a node Load, distributed energy storage device charging, and energy storage device charging; , This represents the total power generation output of photovoltaic and wind power. (2) Power constraint in , , , , , These represent the rated charging and discharging power of distributed energy storage devices, distributed photovoltaic power, distributed wind power, liquefied natural gas devices, hydrogen energy devices, and energy storage devices, respectively. (3) Capacity constraints in , These represent the lower and upper limits of the energy storage system capacity, respectively. (4) Energy conservation constraint This formula indicates that the energy consumption of a distributed energy storage device is less than or equal to the energy storage.

6. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 1, characterized in that, The formula for calculating the error value in S3 is: In the formula, For time period Error value within; Indicates time period The number of nodes in the topology graph; Represents a node During the period The predicted state within the range is 1 for a used state and 0 for an unused state. and Representing nodes respectively During the period The actual and predicted values ​​within.

7. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 6, characterized in that, The formula for calculating the starting node data is: In the formula, The updated starting node data is continuously updated during the optimization process as each node is optimized. Predict data for the starting node.

8. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 5, characterized in that, The specific optimization method for optimizing the energy allocation scheme between the target node and its neighboring nodes in S4 is as follows: In the formula, Represents a node During the period Energy revenue within the area, It is an adjacency matrix. This indicates a node. and nodes There is a connection between them. 0 indicates a node and nodes There is no connection between them or .

9. The multi-energy network scheduling method based on a centralized and distributed dual-layer optimization strategy according to claim 8, characterized in that, The distribution of distributed energy resources in S4 also follows constraints, which are as follows: (1) Flow balance constraint (2) Power constraint (3) Capacity constraints (4) Energy conservation constraint After each node is calculated once, the original... and its adjacent nodes The replacement is the corresponding optimization result.

Images