Comprehensive energy system toughness improving method considering road network traffic uncertainty

Abstract

A comprehensive energy system toughness improvement method considering road network traffic uncertainty belongs to the technical field of power system safety planning operation, and comprises the following steps: constructing a dual uncertainty model containing wind power output uncertainty, photovoltaic output uncertainty and traffic network traffic uncertainty; based on the dual uncertainty model, a two-stage robust optimization method is adopted, pre-disaster resource pre-deployment decision is carried out, and resources comprise fixed electricity-hydrogen-storage resources and mobile hydrogen energy resources; in the disaster occurrence process, on the basis of a pre-disaster resource pre-deployment decision result and system state information obtained in real time, a model prediction control MPC rolling optimization method is adopted, and scheduling and operation strategies of mobile hydrogen energy resources are dynamically adjusted. Through cooperative scheduling of fixed and mobile electricity-hydrogen-storage resources, optimal configuration of hydrogen energy resources before disasters and dynamic response of mobile hydrogen energy in disasters are realized, so that survival and recovery capabilities of the system in extreme weather events are remarkably enhanced.

Description

Technical Field

[0001] This invention belongs to the field of power system safety planning and operation technology, specifically, it relates to a comprehensive energy system resilience enhancement method that takes into account the uncertainty of road network traffic. Background Technology

[0002] Extreme natural disasters, such as snowstorms, typhoons, and earthquakes, have become one of the most severe threats to the safe and stable operation of modern power systems. Therefore, grid resilience has become a new focus in grid planning and operation.

[0003] Currently, the technical means to improve grid resilience mainly revolve around distributed resources, specifically including:

[0004] 1. Islanding Operation Based on Distributed Power Generation: After the main power grid is damaged, local distributed power sources (such as photovoltaic and wind power) are used to form islanded power supply for critical loads. However, the intermittency and strong uncertainty of wind and solar power output are often more pronounced under severe weather conditions, which seriously restricts the reliability of islanded power supply.

[0005] 2. Energy time-shifting using energy storage systems: This involves storing electrical energy when the grid is normal and releasing it during a fault using devices such as stationary energy storage batteries. However, its energy reserves are limited and its geographical location is fixed, making it unable to cope with large-scale, cross-regional power shortages.

[0006] 3. Pre-positioned mobile emergency power supply vehicle: Traditional diesel mobile generators are currently a commonly used emergency measure, but they have inherent limitations such as high operating noise, exhaust emissions that pollute the environment, and a heavy reliance on the fuel supply chain for continuous power supply (fuel supply lines are easily interrupted during disasters).

[0007] Hydrogen energy, as a clean, efficient, and scalable secondary energy carrier, offers a new path to solving the aforementioned challenges. Stationary "electricity-hydrogen-electricity" conversion systems (electrolyzer + hydrogen storage tank + fuel cell) enable cross-seasonal, large-scale energy dispatch and storage. Furthermore, mobile hydrogen storage units and hydrogen fuel cell vehicles with vehicle-to-grid (V2G) capabilities provide unprecedented flexibility in the "spatiotemporal transfer" of hydrogen energy resources.

[0008] However, efficiently integrating mobile hydrogen resources into the emergency power grid dispatch system during disasters faces a crucial and unresolved core challenge: in extreme disaster scenarios, the transportation network itself is highly uncertain. Road damage, bridge collapses, traffic control, and congestion caused by panic or rescue efforts all make the transportation routes and times of mobile hydrogen resources extremely difficult to predict. Most existing studies either ignore the deep coupling between the transportation network and the power grid or simplify the process using deterministic transportation models. This idealized assumption is highly likely to cause the optimized dispatch scheme to completely fail in actual implementation due to mobile resources being "unable to travel" or "unable to move at all," thus missing rescue opportunities and even triggering catastrophic social consequences. Summary of the Invention

[0009] The purpose of this invention is to provide a comprehensive energy system resilience enhancement method to address the dual uncertainties of wind and solar power output and traffic flow under extreme disasters. By coordinating the scheduling of stationary and mobile electricity-hydrogen-storage resources, the method achieves optimized allocation of hydrogen energy resources before a disaster and dynamic response of mobile hydrogen energy during a disaster, thereby significantly enhancing the system's survival and recovery capabilities under extreme weather events.

[0010] The technical solution adopted by the present invention to achieve the above objectives is: a method for improving the resilience of a comprehensive energy system considering the uncertainty of road network traffic, the method comprising the following steps:

[0011] S1: Construct a dual uncertainty model that includes uncertainties in wind power output, photovoltaic power output, and traffic network access.

[0012] S2: Based on the aforementioned dual uncertainty model, a two-stage robust optimization method is employed to make pre-disaster resource deployment decisions. These resources include stationary electricity-hydrogen-storage resources and mobile hydrogen energy resources. It should be noted that stationary electricity-hydrogen-storage resources are existing technologies, referring to comprehensive energy hubs permanently or semi-permanently installed at specific geographical locations, integrating electricity input, hydrogen electrolysis production, hydrogen storage, and downstream applications. Their core function is to achieve cross-temporal and spatial conversion and optimized allocation of energy forms, serving as the core infrastructure for constructing new power systems and hydrogen energy networks.

[0013] S3: During a disaster, based on the pre-disaster resource deployment decision results and real-time system status information, the scheduling and operation strategy of the mobile hydrogen energy resources is dynamically adjusted using the Model Predictive Control (MPC) rolling optimization method.

[0014] Furthermore, in step S1, the dual uncertainty model is described by an uncertainty set, including:

[0015] Uncertainty in wind power output WTUncertainty in photovoltaic power output PV and the uncertain set of traffic network access Ξ traffic The traffic network uncertainty set Ξ traffic Defined as:

[0016]

[0017] Among them, z ij z is a binary variable representing the traffic status of road segment (i,j). ij =1 indicates a road blockage, z ij =0 indicates that the road segment is open to traffic; w ij Γ is the weighting coefficient of road segment (i,j), used to reflect the differences in the impact of different levels of road interruption on the system; Γ is the weighted budget parameter; ε represents the set of all road segments in the traffic network.

[0018] Furthermore, the two-stage robust optimization method in step S2 adopts a min-max-min structure, and its objective function is:

[0019]

[0020] Where x is the pre-disaster deployment decision variable, This indicates that x belongs to the feasible decision set. C inv (x) represents the investment cost associated with the pre-disaster deployment decision variable x; C pre (x) represents the cost of disaster preparedness and maintenance; ξ represents an uncertain scenario, ξ∈Ξ indicates that ξ belongs to the uncertain set Ξ, and Ξ is the uncertain set of wind power output Ξ. WT Uncertainty in photovoltaic power output PV and the uncertain set of traffic network access Ξ traffic Composition: Q(x,ξ) is the second-stage optimal operating cost given the pre-disaster deployment decision variable x and the uncertain scenario ξ, including load reduction penalty.

[0021] Furthermore, step S2 employs the column and constraint generation C&CG algorithm to iteratively solve the two-stage robust optimization model, specifically including:

[0022] S21: Initialize the scene set and set iteration parameters;

[0023] S22: Solve the main problem to obtain the current resource deployment decision plan x;

[0024] S23: For the current deployment decision plan x, solve the sub-problems to find the worst uncertainty scenario ξ that maximizes the system operating cost;

[0025] S24: Add the worst-case uncertainty scenario ξ to the scenario set;

[0026] S25: Repeat steps S22 to S24 until the convergence condition is met, and output the final pre-disaster robust pre-deployment plan.

[0027] Furthermore, the objective function of the MPC rolling optimization method in step S3 is:

[0028]

[0029] Where T represents the total number of time periods in the optimization time domain; represents the system operating cost during time period t; L represents the set of critical loads; s l,t ρ represents the reduction amount of critical load l during time period t; l This represents the reduction penalty weight for critical load l; Φ(x) T ) represents the terminal state penalty function; β represents the terminal weight coefficient.

[0030] Furthermore, step S3 specifically includes:

[0031] S31: Initialize the rolling optimization time t=0 and obtain the current system state information;

[0032] S32: Update the system status and correct the prediction model based on the latest measurement data;

[0033] S33: At the current time t, based on the updated state and the prediction of future time periods, solve the finite-time optimization problem to obtain the optimal scheduling sequence;

[0034] S34: Execute the scheduling decision corresponding to the current time in the optimal scheduling sequence;

[0035] S35: Wait for the next moment, obtain new system state information, let t = t + 1, repeat steps S32 to S35 until the optimization time domain endpoint is reached.

[0036] Furthermore, the mobile hydrogen energy resources include hydrogen fuel cell vehicles equipped with vehicle-to-grid (V2G) interaction capabilities.

[0037] Furthermore, the stationary electricity-hydrogen-storage resource includes an electricity-hydrogen-electricity conversion system consisting of an electrolyzer, a hydrogen storage tank, and a fuel cell.

[0038] The beneficial effects of the present invention through the above design scheme are as follows:

[0039] Constructing a dual uncertainty model to enhance decision robustness: Innovatively incorporating the uncertainty of traffic network status and the fluctuation of wind and solar power output into the uncertainty set, and adopting a two-stage robust optimization method, the pre-disaster scheduling decision can withstand a wider range of risk scenarios, significantly improving the system's ability to predict and its robustness in response to extreme disasters.

[0040] Achieving synergy between fixed and mobile resources and innovating resilience enhancement methods: Breaking away from the traditional model of relying on fixed energy storage to enhance resilience, it pioneered the use of hydrogen fuel cell vehicles as a flexible resource for dispatch. Through cross-domain coupling and synergistic interaction of electricity, hydrogen, and storage resources, a three-dimensional resilience support system of "fixed deployment and mobile reinforcement" has been formed.

[0041] The system is optimized in two phases before and during disasters to ensure its economic efficiency and stability: Before a disaster, robust pre-disaster scheduling enables the strategic deployment of key resources such as hydrogen energy, laying a foundation for resilience while ensuring economic efficiency; during a disaster, rolling optimization is carried out in conjunction with real-time road condition information to dynamically adjust the dispatch path and operation strategy of mobile energy storage, accurately support critical loads, effectively suppress the spread of disasters, and ensure the stable operation of the system's core functions.

[0042] Fully leverage the multiple advantages of hydrogen energy to create an efficient and resilient carrier: Using hydrogen energy as the core medium for enhancing resilience, leveraging its characteristics of long-term energy storage, ease of transportation, and clean and efficient operation, hydrogen fuel cell vehicles can directly deliver energy to key nodes, realizing the spatial and temporal transfer and precise delivery of energy, greatly improving the flexibility, reliability, and overall resilience of energy supply. Attached Figure Description

[0043] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and are used to understand the invention, but do not constitute an improper limitation of the invention. In the drawings:

[0044] Figure 1 A schematic diagram of a two-stage robust and rolling optimization model framework;

[0045] Figure 2 This is a flowchart of the two-stage solution process for the two-stage robust optimization method in this embodiment of the invention. Detailed Implementation

[0046] To more clearly illustrate the present invention, the following description, in conjunction with preferred embodiments and accompanying drawings, further clarifies the invention. Those skilled in the art should understand that the specific description below is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. To avoid obscuring the essence of the invention, well-known methods, processes, flows, elements, and circuits are not described in detail.

[0047] This invention proposes a comprehensive energy system resilience enhancement method considering road network traffic uncertainty. This method employs a two-stage robust and rolling optimization model framework, mainly consisting of a pre-disaster preparation stage and a disaster implementation stage. During the pre-disaster preparation stage, wind and solar power output, load demand, and traffic conditions are predicted at the time of the disaster. The worst-case scenario (no wind and solar power output, significant power outages, and widespread traffic paralysis) is taken into consideration. Fixed energy storage, mobile energy storage, and hydrogen fuel cell vehicles are deployed in advance for this scenario. A robust optimization model is established with the goal of minimizing costs, and the optimal result is obtained using C&CG (column and constraint generation). In the disaster implementation stage, pre-disaster decision data is used as initial data. A rolling optimization model is established with the goal of minimizing the total operating cost and load reduction penalty within the current prediction time domain. This model is then solved using Mixed Integer Programming (MIP). Combined with real-time road conditions and energy status, fixed electricity-hydrogen-storage resources and mobile hydrogen energy resources are coordinated and scheduled to minimize the total system cost and enhance system resilience.

[0048] Specifically:

[0049] (1) Robust optimization

[0050] A two-stage robust approach is adopted to handle the uncertainties of wind power, photovoltaic power, and road network access. The established model has a min-max-min structure, and the optimization objective is to minimize the cost in the worst-case scenario. The system is modeled as a two-stage optimization model, where the first stage is the decision-making strategy layer (initial deployment location and number of hydrogen fuel cell vehicles and mobile energy storage vehicles, as well as investment in fixed equipment, etc.), and the second stage is the operation scheduling decision-making strategy layer (schedule of mobile resources, charging / discharging / generating power, load reduction, etc.).

[0051] Construction of indeterminate sets:

[0052]

[0053] Among them, Ξ traffic Represents an uncertain set of the transportation network; Ξ WT Represents the uncertain set of wind power output; Ξ PV Represents the uncertain set of photovoltaic output; z ij ∈{0,1} represents the blocking indication for road segment (i,j), where 1 indicates blocking and 0 indicates passage; weight w ij Based on road category settings, for example, you can select national highway. ij =3, Provincial Highway w ij =2, rural road w ij=1. These weights reflect that "the impact of blocking a highway on the system is greater than blocking a side road". Γ is the weighted budget (for example, Γ=6 means that the blockable combination is equivalent to 6 rural roads or 2 national highways, and so on). and P represents the actual values ​​of wind power and solar power during time period t; WT,t and P PV,t τ represents the predicted values ​​of wind power and solar power respectively during time period t; WT,t and τ PV,t These represent the fluctuation states of wind power and photovoltaic power during time period t, respectively, with values ​​of 0 or 1; WT,t and l PV,t Ω represents the fluctuation values ​​of wind power and photovoltaic power during time period t, respectively; WT and Ω PV This represents the uncertainty of wind and solar power, indicating the number of periods within a cycle where wind and solar power output fluctuates, z∈Ξ traffic ε is a weighted blocking variable, representing the set of all road segments in the traffic network.

[0054] Objective function:

[0055] Phase 1: Deployment Decisions

[0056]

[0057] Where ξ=(ξ WT ,ξ PV ,z), where z∈Ξ traffic ξ is a weighted blocking variable; WT ∈Ξ WT ξ is the weighted photovoltaic power output variable; PV ∈Ξ PV C represents the weighted wind turbine output variable; Q(x,ξ) is the optimal disaster recovery operating cost (including load shedding penalty) for a given deployment x under scenario ξ. inv (x): Investment cost for deploying / purchasing one unit of resource. C pre (x): Costs such as disaster preparedness / maintenance. For the set of all deployment schemes,

[0058] Subproblems (for a given x) ★ Find the worst-case scenario: The scenery is unhelpful, and the predicted road is completely impassable. It should be noted that x and x... ★ They are different; x contains x. ★ (Worst-case deployment decision).

[0059]

[0060] Adversarial outer layer (find worst-case scenario):

[0061]

[0062] Where C oper (u) represents operating costs (fuel / electricity / hydrogen consumption costs, vehicle operating costs, etc.); ρ l For load reduction penalty; s l For reduction amount; Given x ★ The set of feasible operational strategies for (deployment) and scenario ξ; Q(x) ★ ,ξ) is the worst-case deployment x given in scenario ξ. ★ The optimal disaster recovery operating cost (including load reduction penalties); u represents the operational decision scheme; l represents the critical load.

[0063] Solution process:

[0064] Step 1: Initialize the scenario set S, which contains all possible worst-case scenarios;

[0065] Step 2: Solve the main problem to obtain the resource deployment decision x. ★ ;

[0066] Step 3: Based on x ★ Solve the subproblem (maximize the cost of the worst-case scenario);

[0067] Step 4: Consider the worst-case scenario. ★ Add to scene set S;

[0068] Step 5: Iterate until the main problem and subproblems converge.

[0069] (2) Second phase MPC rolling

[0070] MPC objective function:

[0071]

[0072] in, The operating costs at time t (including fuel / hydrogen consumption costs, mobile vehicle operating costs, charging costs, etc.); s l,t ≥0 represents the reduction amount of the critical load l at time t, ρ l Its penalty weight; Φ(x) T ) represents the terminal cost / state penalty, β represents the terminal weight, T represents the total number of time periods in the optimization time domain, and L represents the set of critical loads;

[0073] Solution process:

[0074] Step 1: Obtain the latest real-time information (such as traffic conditions, wind and solar power output, etc.).

[0075] Step 2: Use predictive models to optimize the scheduling of the next control step.

[0076] Step 3: Execute the current scheduling strategy (such as scheduling certain vehicles or adjusting the charging and discharging strategy).

[0077] Step 4: Scroll the control window (t0→t) 0+1 ), and recalculate the next scheduling step.

[0078] Step 5: Iterate until the time window T is reached.

[0079] The specific two-stage solution process is as follows:

[0080] Phase 1: Pre-disaster Robust Deployment

[0081] Objective: To determine the initial deployment plan for mobile energy storage and hydrogen fuel cell vehicle resources to cope with the worst-case disaster scenarios.

[0082] Step 1: Input the system's fixed parameters and uncertainties. Fixed parameters include network topology, load, and distributed generation parameters; uncertainties describe potential failure scenarios caused by disasters (e.g., branch line interruptions, fluctuations in photovoltaic output, etc.). ("etc." are uncertain terms and must be exhaustively listed.)

[0083] Step 2: Initialization. Set the iteration count k = 0, upper bound UB = +∞, lower bound LB = -∞, and convergence threshold l.

[0084] Step 3: Solve the master problem. The master problem is a mixed-integer linear programming (MILP) problem, the goal of which is to minimize investment costs (such as fixed energy storage deployment costs) and expected operating costs, while satisfying operational constraints under all possible failure scenarios. The master problem gives the current deployment scheme x_k.

[0085] Step 4: Solve the subproblem. The subproblem, given the current deployment scheme x_k, seeks the worst-case failure scenario u_k (i.e., the scenario that maximizes the operating cost). The subproblem is a two-level optimization problem: the inner level optimizes the runtime schedule under given failure scenarios, while the outer level seeks the failure scenario that maximizes the operating cost. Typically, the subproblem is transformed into a mixed-integer linear programming problem.

[0086] Step 5: Update upper and lower bounds. Calculate the total cost of the current deployment plan in the worst-case scenario, and update the upper bound UB = min(UB, total cost). Meanwhile, the objective value given by the main problem is the lower bound LB.

[0087] Step 6: Convergence check. If (UB-LB) / UB≤l, the algorithm converges, and the current deployment scheme x_k is output as a robust pre-deployment scheme; otherwise, the worst-case scenario u_k is added as a new scenario to the scenario set of the main problem, k = k+1, and the algorithm returns to step 3.

[0088] Note: The first stage uses the column and constraint generation (C&CG) algorithm, which iteratively adds worst-case scenarios to approximate the robust optimization solution.

[0089] Phase Two: Disaster-Related MPC Rolling Optimization

[0090] Objective: During a disaster, adjust the scheduling of mobile energy storage and hydrogen fuel cell vehicle resources in real time based on the actual failure situation to minimize operating costs and meet safety constraints.

[0091] Step 1: Initialize the rolling time t=0. Receive the current system status (such as load, photovoltaic output, wind turbine output, road conditions, and network topology).

[0092] Step 2: Feedback Correction. Based on the actual measurement data, update the current system state and correct the prediction model.

[0093] Step 3: Solve the MPC problem. At the current time t, based on the current state and the predicted load, photovoltaic output, and wind turbine output for the next H time periods, solve a mixed integer linear programming (MILP) problem to optimize scheduling decisions for the next H time periods (such as the charging and discharging of mobile energy storage, and the scheduling of hydrogen fuel cell vehicles and mobile energy storage). The objective of this problem is to minimize the total operating cost for the next H time periods while satisfying system operating constraints.

[0094] Step 4: Execute the decision. Apply only the decision obtained from the first time interval of the MPC problem solution (i.e., the actions from time t to t+1) to advance the system state to time t+1.

[0095] Step 5: Wait for new measurement data. At time t+1, obtain the new system status (such as actual fault conditions, load changes, changes in power output from chemical and wind power plants, and road conditions).

[0096] Step 6: Check if the optimization time domain endpoint has been reached. If yes, end; otherwise, let t = t + 1 and return to step 2.

[0097] The entire algorithm works as follows: First, a robust pre-deployment plan is obtained through the first stage. Then, during the disaster, MPC rolling optimization is used for real-time adjustments. This ensures the robustness of the pre-deployment plan to the worst-case scenario and also improves the system's resilience by using real-time information for optimized scheduling during actual disasters.

[0098] To more specifically illustrate the implementation process and technical effects of the present invention, the following example, using a virtual but typical integrated energy system in a coastal city, provides a specific application instance of this method.

[0099] Application Example: Enhancing the Resilience of a Coastal City's Power Grid in the Face of Typhoon Disasters

[0100] 1. System Overview

[0101] Location: A coastal city, vulnerable to typhoons.

[0102] Power network: Includes three 110kV substations (S1, S2, S3), supplying power to the city's core area, industrial zone, and emergency command center. Critical loads include: the city hospital (L1), the emergency command center (L2), and the communication hub (L3).

[0103] Distributed power generation: Wind farms (WT) are located in the northern part of the city, while photovoltaic (PV) power stations are distributed in the suburbs and on building rooftops.

[0104] Stationary electricity-hydrogen-storage resources: An electricity-hydrogen-electricity conversion system is configured at the city energy hub (node ​​E1), including a 2MW electrolyzer, a 500kg hydrogen storage tank, and a 1MW fuel cell.

[0105] Mobile hydrogen energy resources: It has 10 hydrogen fuel cell vehicles (HFCVs) with V2G capabilities, each with a hydrogen storage capacity of about 5kg and a maximum discharge power of 50kW.

[0106] Transportation network: The main roads include two national highways (G1 and G2), three provincial highways (P1, P2, and P3), and several rural roads. The connectivity of the road network directly affects the scheduling efficiency of HFCV.

[0107] 2. Disaster Setting

[0108] The typhoon is expected to make landfall in 24 hours, and may cause:

[0109] Wind and solar power output has dropped significantly or even to zero;

[0110] Some power transmission lines were interrupted;

[0111] Roads are damaged (especially along the coast and in low-lying areas), and traffic capacity is uncertain.

[0112] 3. Disaster preparedness phase (using two-stage robust optimization)

[0113] Step S1: Construct a dual uncertainty model

[0114] Uncertainty in wind power output WT The wind power output is set to zero for a maximum of four periods under the influence of a typhoon.

[0115] Uncertainty in photovoltaic output PV Set a maximum of 6 time periods during which the photovoltaic output is zero.

[0116] Uncertainty set of traffic network access Ξtraffic :

[0117] Weight setting: National Highway w ij =3, Provincial Highway w ij =2, rural road w ij =1.

[0118] Set the budget Γ = 8, which means that the expected "weighted total severity" of road network disruptions will not exceed 8 (for example: 2 national highways are disrupted + 1 rural road is disrupted, i.e. 3×2+1=7≤8).

[0119] Step S2: Two-stage robust optimization decision

[0120] Main issue (investment and deployment decisions):

[0121] The decision variables x include: whether to add fixed hydrogen storage tanks and the initial deployment location of HFCVs (e.g., how many vehicles to allocate to the vicinity of E1 and S1).

[0122] Objective: Minimize investment and upfront costs, plus operating costs to handle worst-case scenarios.

[0123] Sub-problem (finding the worst-case scenario):

[0124] Given a deployment scheme x, find the combination of wind and solar power output and road interruption that maximizes the system's operating cost.

[0125] For example, if wind and solar power are completely shut down and G1 and P1 are simultaneously interrupted (with a weighted sum of 5), HFCV will be unable to be dispatched from energy hub E1 to critical load L2.

[0126] C&CG iterative solution:

[0127] After five rounds of iteration, the robust pre-deployment scheme was obtained:

[0128] Three HFCVs were pre-positioned near S1;

[0129] Add a 200kg hydrogen storage tank to E1;

[0130] Keep the remaining HFCVs on standby in the central garage.

[0131] 4. Disaster Implementation Phase (using MPC rolling optimization)

[0132] Actual disaster situation after the typhoon made landfall:

[0133] Wind power was completely shut down, and solar power output dropped to 10%.

[0134] Transmission lines S1-S2 are interrupted;

[0135] National Highway G1 is partially closed due to fallen trees (passable slowly), and Provincial Highway P1 is severely flooded (completely closed).

[0136] Step S3: MPC Rolling Optimization Scheduling

[0137] Rolling time domain: T = 4 hours, rolling once every 30 minutes.

[0138] Current time t=0:

[0139] System status: L1 is powered normally by S1; L2 is out of power due to line interruption; L3 is supported by residual photovoltaic voltage.

[0140] Road condition update: G1 is passable but speed is limited; P1 is closed.

[0141] MPC optimization solution (t=0 to t=4):

[0142] Objective: Minimize operating costs plus load reduction penalties.

[0143] Decision: Schedule E1's HFCV to bypass G1 to support L2; Schedule S1's pre-configured HFCV to provide backup power for L3.

[0144] Execute the current decision:

[0145] Two HFCVs departed from E1 and are expected to arrive at L2 in 60 minutes.

[0146] One HFCV from S1 starts up, providing 30kW of support for L3.

[0147] Update the state at t=1:

[0148] L2 power supply restored (HFCV arrives);

[0149] New traffic conditions: Traffic on G1 has improved, but congestion has occurred on P2.

[0150] The system was re-optimized and the routing and power scheduling of subsequent vehicles were adjusted.

[0151] 5. Comparison of Implementation Results

[0152] Traditional methods (relying solely on stationary energy storage and diesel generators):

[0153] Diesel vehicles are unable to reach L2 due to road closures;

[0154] The stationary energy storage capacity is depleted quickly, and L2 and L3 experienced continuous power outages for more than 3 hours.

[0155] Method of the present invention:

[0156] Pre-disaster HFCV rapid response;

[0157] Real-time traffic updates and dynamic route adjustments ensure energy delivery; L2 power outage time was reduced to 1 hour, and no L3 power outage occurred.

[0158] The total system load reduction was reduced, and the resilience was significantly improved.

Claims

1. A method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty, characterized in that, The method includes the following steps: S1: Construct a dual uncertainty model that includes uncertainties in wind power output, photovoltaic power output, and traffic network access. S2: Based on the aforementioned dual uncertainty model, a two-stage robust optimization method is adopted to make pre-disaster resource deployment decisions, wherein the resources include stationary electricity-hydrogen-storage resources and mobile hydrogen energy resources; S3: During a disaster, based on the pre-disaster resource deployment decision results and real-time system status information, the scheduling and operation strategy of the mobile hydrogen energy resources is dynamically adjusted using the Model Predictive Control (MPC) rolling optimization method.

2. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 1, characterized in that, In step S1, the dual uncertainty model is described by an uncertainty set, including: Uncertainty in wind power output WT Uncertainty in photovoltaic power output PV and the uncertain set of traffic network access Ξ traffic The traffic network uncertainty set Ξ traffic Defined as: Among them, zi j Zi is a binary variable representing the traffic status of road segment (i,j). j =1 indicates a road blockage, zi j =0 indicates that the road segment is open to traffic; wi j Γ is the weighting coefficient of road segment (i,j), used to reflect the differences in the impact of different levels of road interruption on the system; Γ is the weighted budget parameter; ε represents the set of all road segments in the traffic network.

3. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 2, characterized in that, The two-stage robust optimization method in step S2 adopts a min-max-min structure, and its objective function is: Where x is the pre-disaster deployment decision variable, and x∈χ means that x belongs to the feasible decision set χ; C inv (x) represents the investment cost associated with the pre-disaster deployment decision variable x; Cpre(x) represents the pre-disaster preparation and maintenance costs; ξ represents an uncertain scenario, ξ∈Ξ indicates that ξ belongs to the uncertain set Ξ, and Ξ is composed of the uncertain set Ξ of wind power output. WT Uncertainty in photovoltaic power output PV and the uncertain set of traffic network access Ξ traffic The composition is as follows: Q(x,ξ) is the second-stage optimal operating cost given the pre-disaster deployment decision variable x and the uncertain scenario ξ, including the load reduction penalty.

4. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 3, characterized in that, Step S2 employs the column and constraint generation C&CG algorithm to iteratively solve the two-stage robust optimization model, specifically including: S21: Initialize the scene set and set iteration parameters; S22: Solve the main problem to obtain the worst-case deployment decision x for the current resources. ＊ ; S23: For x ＊ Solve the subproblems to find the worst-case uncertainty scenario ξ that maximizes the system's operating cost. ＊ ; S24: The worst uncertainty scenario ξ ★ Add to the aforementioned scene set; S25: Repeat steps S22 to S24 until the convergence condition is met, and output the final pre-disaster robust pre-deployment plan.

5. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 1, characterized in that, The objective function of the MPC rolling optimization method in step S3 is: Where T represents the total number of time periods in the optimization time domain; represents the system operating cost during time period t; L represents the set of critical loads; s l,t ρ represents the reduction amount of critical load l during time period t; l This represents the reduction penalty weight for critical load l; Φ(x) T ) represents the terminal state penalty function; β represents the terminal weight coefficient.

6. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 5, characterized in that, Step S3 specifically includes: S31: Initialize the rolling optimization time t=0 and obtain the current system state information; S32: Update the system status and correct the prediction model based on the latest measurement data; S33: At the current time t, based on the updated state and the prediction of future time periods, solve the finite-time optimization problem to obtain the optimal scheduling sequence; S34: Execute the scheduling decision corresponding to the current time in the optimal scheduling sequence; S35: Wait for the next moment, obtain new system state information, let t = t + 1, repeat steps S32 to S35 until the optimization time domain endpoint is reached.

7. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 1, characterized in that, The mobile hydrogen energy resources include hydrogen fuel cell vehicles equipped with vehicle-to-grid (V2G) functionality.

8. The method for enhancing the resilience of a comprehensive energy system considering road network traffic uncertainty according to claim 1, characterized in that, The stationary electro-hydrogen-storage resource includes an electro-hydrogen-electric conversion system consisting of an electrolyzer, a hydrogen storage tank, and a fuel cell.

Images