Security rule extraction and optimization method for improving urban energy system resilience

Abstract

The present application relates to a method for extracting and optimizing security rules for improving the resilience of urban energy systems, including: calculating the resilience index of urban energy systems and constructing a two-layer optimization model of urban energy systems; modeling the dynamic characteristics of the power electronics link of urban energy systems; modeling the dynamic characteristics of multi-energy networks of urban energy systems; based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, constructing a two-stage resilience operation model of the urban energy system under disaster events, and based on a preset sparse weight oblique decision tree model, converting the two-stage resilience model of the urban energy system under disaster events into mixed integer linear constraints, and embedding it into the two-layer optimization model of the urban energy system to obtain the optimized operation simulation results under the mixed integer linear constraints. Thus, a security rule with a simple form, strong interpretability and strong operability is provided, which can significantly improve the solution efficiency, has significant innovation and good application value.

Description

Technical Field

[0001] The present application relates to the field of energy scheduling technology, and in particular to a method for extracting and optimizing security rules for improving the resilience of urban energy systems. Background Art

[0002] Extreme natural disasters such as hurricanes, earthquakes, snowstorms, and high temperatures have had a devastating impact on energy systems around the world, posing a severe challenge to the secure energy supply for large-scale urban users. Therefore, urban energy systems must not only be highly reliable to cope with small-scale, high-probability, everyday failures, but also highly resilient to cope with large-scale, low-probability, extreme disasters.

[0003] In recent years, urban energy systems have faced two major new trends: first, the significant increase in the coupling of different energy sources; second, the continuous advancement of power electronics. Specifically, multiple energy sources—electricity, gas, heat, and cooling—are interconnected at multiple stages, including production, transmission, conversion, storage, and utilization. Compared to discrete energy systems (such as the power system), urban energy systems are expected to further enhance system resilience in the following three areas: 1) Energy conversion: The integration of diverse energy conversion equipment (such as energy hubs, gas turbines, power-to-gas, and electric boilers) enables cross-system energy transfer, significantly improving energy supply flexibility. 2) Energy storage: The coordination of centralized energy storage (such as compressed air energy storage and pumped hydro), distributed energy storage (such as electric vehicles), and generalized energy storage (such as energy stored in gas / heat / cooling pipeline networks) can provide sufficient multi-energy backup. 3) Energy utilization: The thermal inertia of heating / cooling loads (such as space heating / cooling loads) ensures that the failure of a heating / cooling source will not significantly affect user comfort in the short term, thus relaxing the constraints on the instantaneous balance of supply and demand to a certain extent. On the other hand, power electronic equipment has been widely used in all links of the urban energy system, such as photovoltaic and wind power conversion equipment in the production link, AC / DC conversion equipment and flexible soft switches (soft open points, SOP) in the transmission link, heat pumps and compression refrigerators in the conversion link, batteries and electric vehicles in the storage link, and emerging DC loads (such as data centers) in the utilization link.

[0004] Although relevant research on improving the resilience of urban energy systems in related technologies has made certain progress in models and methodologies, some issues still need further exploration. From a model perspective, most research on related technologies has overlooked the emergency supply capacity of heat / cold pipe networks in disaster events and the buffering role of heat / cold loads in disaster events. On the other hand, the rapid action and short-term overload capacity of massive power electronic equipment have not been fully exploited, which can provide short-term support for the energy supply of important loads after disaster events. From a methodological perspective, related technologies are mainly based on model-driven methods, embedding the system operation model under disaster events into the upper-level system optimization operation problem through methods such as stochastic optimization and robust optimization. However, the complex energy conversion structure of urban energy systems and the wide variety of disaster event scenarios will lead to a large optimization problem scale and high solution complexity. Summary of the Invention

[0005] The present application provides a safety rule extraction and optimization method, device, electronic device and storage medium for improving the resilience of urban energy systems, in order to solve the problem that related technologies, on the one hand, ignore the emergency supply capacity and buffering effect of hot / cold pipelines in disaster events, and on the other hand, lead to large-scale optimization problems and high solution complexity. It provides rules with simple form, strong interpretability, strong operability and safety, which can significantly improve the solution efficiency and have significant innovation and good application value.

[0006] The first aspect of the present application provides a method for extracting and optimizing safety rules for improving the resilience of urban energy systems, comprising the following steps: calculating the resilience index of the urban energy system, and constructing a two-layer optimization model of the urban energy system based on the resilience index; modeling the dynamic characteristics of the power electronics link of the urban energy system based on a preset energy transmission link, a preset energy storage link and a preset energy conversion link to obtain a dynamic characteristics model of the power electronics link; modeling the dynamic characteristics of the multi-energy network of the urban energy system based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipeline to obtain a dynamic characteristics model of the multi-energy network; and constructing a two-stage resilience operation model of the urban energy system under disaster events based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, and based on a preset sparse weight oblique decision tree model, converting the two-stage resilience model of the urban energy system under the disaster event into a mixed integer linear constraint, and embedding the mixed integer linear constraint into the two-layer optimization model of the urban energy system to obtain an optimization operation simulation result under the mixed integer linear constraint.

[0007] Optionally, in some embodiments, calculating the urban energy system resilience index includes: calculating the urban energy system resilience index based on a preset resilience index formula, wherein the preset resilience index formula is:

[0008]

[0009] Among them, ε R is the resilience index, SP0 is the system performance before the disaster event, SP d is the minimum value of system performance after a disaster event, P load is the system load.

[0010] Optionally, in some embodiments, before modeling the dynamic characteristics of the power electronics link of the urban energy system based on the preset energy transmission link, the preset energy storage link, and the preset energy conversion link to obtain the dynamic characteristics model of the power electronics link, the method further includes: determining the preset energy transmission link based on a first instantaneous adjustment strategy of a response speed of a flexible soft switch, wherein the first instantaneous adjustment strategy is:

[0011] P SOP (t)≤M·(1-S line (t));

[0012] 0≤P SOP (t)≤P SOP,max ;

[0013] in, is the important load of the system, P SOP () is the operating power of the flexible soft switch at time t, M is a positive number, S line is the operating status of the distribution line, P SOP,max The upper limit of the operating power of the flexible soft switch;

[0014] The preset energy storage link is determined based on a second instantaneous regulation strategy of the discharge power of the battery, wherein the second instantaneous regulation strategy is:

[0015] 0≤P ES (t0)≤P ES,max ;

[0016] Among them, P ES is the operating power of the battery, P ES, is the upper limit of the battery's operating power;

[0017] The preset energy conversion link is determined based on a third instantaneous adjustment strategy of the input electrical power of the heat pump and the input electrical power of the compression refrigerator, wherein the third instantaneous adjustment strategy is:

[0018] 0≤P EHP (t0)≤P EHP,max ;

[0019] 0≤P CERG (t0)≤P CERG,max ;

[0020] Among them, P EHP is the heat pump operating power, P EHP,max is the upper limit of the heat pump operating power, P CERG is the operating power of the compression refrigerator, P CERG, It is the upper limit of the operating power of the compression refrigerator.

[0021] Optionally, in some embodiments, the multi-energy network dynamic characteristics of the urban energy system are modeled based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network, and the storage capacity of the regional heating / cooling network pipelines to obtain a multi-energy network dynamic characteristics model, including: determining a first relationship between the energy reserves of the pipeline network equivalent energy storage at the moment of the disaster event and the operating status of the heat source at the moment of the disaster event based on the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipelines; determining a second relationship between the energy reserves of the load thermal inertia equivalent energy storage at the moment of the disaster event and the heat / cooling load at the moment of the disaster event based on the real-time balance of supply and demand of the power system; and modeling the multi-energy network dynamic characteristics of the urban energy system according to the first relationship and the second relationship to obtain a multi-energy network dynamic characteristics model.

[0022] Optionally, in some embodiments, the objective function of the two-stage resilience operation model of the urban energy system is:

[0023]

[0024] Where k is the load, LS k is the load loss size of load k, Ω L is the load set, and t is the time.

[0025] Optionally, in some embodiments, the two-stage resilience model of the urban energy system under the disaster event is converted into a mixed integer linear constraint based on the preset sparse weight oblique decision tree model, including: obtaining a sparse weight oblique decision tree based on the preset sparse weight oblique decision tree model; using the optimization objective function of each node of the sparse weight oblique decision tree to convert the two-stage resilience model of the urban energy system under the disaster event into a mixed integer linear constraint.

[0026] Optionally, in some embodiments, the optimization objective function of the sparse weighted oblique decision tree at each node is:

[0027]

[0028] Among them, θ is the optimal partition parameter vector, N is the number of samples, W L (θ) is the weight sum of the left child node, W R (θ) is the weight sum of the right child node, H L (θ) is the weighted information entropy of the left child node, H R (θ) is the weighted information entropy of the right child node, λ1|θ|+λ2‖θ‖ 2 is the Elastic Net regularization term.

[0029] The second aspect of the present application provides a safety rule extraction and optimization device for improving the resilience of urban energy systems, including: a calculation module for calculating the resilience index of the urban energy system and constructing a two-layer optimization model of the urban energy system based on the resilience index; a generation module for modeling the dynamic characteristics of the power electronics link of the urban energy system based on a preset energy transmission link, a preset energy storage link and a preset energy conversion link, and obtaining a dynamic characteristics model of the power electronics link; a modeling module for modeling the dynamic characteristics of the multi-energy network of the urban energy system based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipeline, and obtaining a dynamic characteristics model of the multi-energy network; and an optimization module for constructing a two-stage resilience operation model of the urban energy system under disaster events based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, and based on a preset sparse weight oblique decision tree model, converting the two-stage resilience model of the urban energy system under the disaster event into a mixed integer linear constraint, and embedding the mixed integer linear constraint into the two-layer optimization model of the urban energy system to obtain the optimization operation simulation result under the mixed integer linear constraint.

[0030] Optionally, in some embodiments, the calculation module is further configured to calculate the urban energy system resilience index based on a preset resilience index formula, wherein the preset resilience index formula is:

[0031]

[0032] Among them, ε R is the resilience index, SP0 is the system performance before the disaster event, SP d is the minimum value of system performance after a disaster event, P load is the system load.

[0033] Optionally, in some embodiments, the generating module is further configured to: determine the preset energy transmission link based on a first instantaneous adjustment strategy of a response speed of the flexible soft switch, wherein the first instantaneous adjustment strategy is:

[0034] P SOP (t)≤M·(1-S line (t));

[0035] 0≤P SOP (t)≤P SOP,max ;

[0036] in, is the important load of the system, P SOP (t) is the operating power of the flexible soft switch at time t, M is a positive number, S line is the operating status of the distribution line, P SOP,max The upper limit of the operating power of the flexible soft switch;

[0037] The preset energy storage link is determined based on a second instantaneous regulation strategy of the discharge power of the battery, wherein the second instantaneous regulation strategy is:

[0038] 0≤P ES (t0)≤P ES,max ;

[0039] Among them, P ES is the operating power of the battery, P ES,max is the upper limit of the battery's operating power;

[0040] The preset energy conversion link is determined based on a third instantaneous adjustment strategy of the input electrical power of the heat pump and the input electrical power of the compression refrigerator, wherein the third instantaneous adjustment strategy is:

[0041] 0≤P EHP (t0)≤P EHP,max ;

[0042] 0≤P CERG (t0)≤P CERG,max ;

[0043] Among them, P EHP is the heat pump operating power, P EHP,max is the upper limit of the heat pump operating power, P CERG is the operating power of the compression refrigerator, P CERG,max It is the upper limit of the operating power of the compression refrigerator.

[0044] Optionally, in some embodiments, the modeling module is further used to: determine a first relationship between the energy reserves of the pipeline network equivalent energy storage at the time of the disaster event and the operating status of the heat source at the time of the disaster event based on the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipeline; determine a second relationship between the energy reserves of the load thermal inertia equivalent energy storage at the time of the disaster event and the heat / cooling load at the time of the disaster event based on the real-time balance of supply and demand of the power system; model the multi-energy network dynamic characteristics of the urban energy system according to the first relationship and the second relationship to obtain a multi-energy network dynamic characteristics model.

[0045] Optionally, in some embodiments, the objective function of the two-stage resilience operation model of the urban energy system is:

[0046]

[0047] Where k is the load, LS k is the load loss size of load k, Ω L is the load set, and t is the time.

[0048] Optionally, in some embodiments, the optimization module is also used to: obtain a sparse weight oblique decision tree based on the preset sparse weight oblique decision tree model; and use the optimization objective function of each node of the sparse weight oblique decision tree to convert the two-stage resilience model of the urban energy system under the disaster event into a mixed integer linear constraint.

[0049] Optionally, in some embodiments, the optimization objective function of the sparse weighted oblique decision tree at each node is:

[0050]

[0051] Among them, θ is the optimal partition parameter vector, N is the number of samples, W L (θ) is the weight sum of the left child node, W R (θ) is the weight sum of the right child node, H L (θ) is the weighted information entropy of the left child node, H R (θ) is the weighted information entropy of the right child node, λ1|θ|+λ2‖θ‖ 2 is the Elastic Net regularization term.

[0052] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the security rule extraction and optimization method for improving the resilience of urban energy systems as described in the above embodiments.

[0053] The fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the security rule extraction and optimization method for improving the resilience of urban energy systems as described in the above embodiments.

[0054] Therefore, by coordinating the resilience resources of power electronics at two time scales, the fast dynamics of power electronics and the slow dynamics of multi-energy networks, a two-stage resilient operation model for urban energy systems under disaster events was proposed, effectively ensuring the continuous supply of critical loads during disaster events. Furthermore, a data-driven method for extracting and embedding security rules for urban energy systems in optimized operation was proposed. This method transforms the complex system resilience model into a small number of mixed-integer linear constraints, namely security rules, which are embedded into the system's normal operation model. This significantly improves the efficiency of solving optimization problems and has high application value.

[0055] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0057] Figure 1 A flowchart of a method for extracting and optimizing security rules for improving the resilience of urban energy systems according to an embodiment of the present application;

[0058] Figure 2 A flowchart of a method for extracting and optimizing security rules for improving the resilience of urban energy systems according to a specific embodiment of the present application;

[0059] Figure 3 A schematic diagram of a system topology structure provided according to a specific embodiment of the present application;

[0060] Figure 4 A schematic diagram of the annual electricity, heating, and cooling load curves of the system provided according to a specific embodiment of the present application;

[0061] Figure 5 This is a schematic diagram of annual photovoltaic power generation data of a system provided according to a specific embodiment of the present application;

[0062] Figure 6 A schematic diagram of system operation safety rules (summer) provided according to a specific embodiment of the present application;

[0063] Figure 7 A schematic diagram of the system's safe operating status according to a specific embodiment of the present application;

[0064] Figure 8 This is a schematic diagram showing the comparison results of system operation cost and calculation time provided according to a specific embodiment of the present application;

[0065] Figure 9 Schematic diagram of a block diagram of a security rule extraction and optimization device for improving the resilience of an urban energy system according to an embodiment of the present application;

[0066] Figure 10 A schematic diagram of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0067] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0068] The following describes, with reference to the accompanying drawings, the security rule extraction and optimization method, device, electronic device, and storage medium for improving the resilience of urban energy systems in accordance with an embodiment of the present application. In response to the problems mentioned in the above background technology, on the one hand, the relevant technologies ignore the emergency supply capacity and buffering effect of the hot / cold pipeline network under disaster events, and on the other hand, the optimization problem is large in scale and the solution complexity is high. The present application provides a safety rule extraction and optimization method for improving the resilience of urban energy systems. In this method, by calculating the resilience index of the urban energy system and constructing a two-layer optimization model of the urban energy system based on the resilience index, the dynamic characteristics of the power electronics link of the urban energy system are modeled based on the preset energy transmission link, the preset energy storage link and the preset energy conversion link to obtain the dynamic characteristics model of the power electronics link. Based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cold network pipeline, the dynamic characteristics of the multi-energy network of the urban energy system are modeled to obtain the dynamic characteristics model of the multi-energy network. Based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, a two-stage resilience operation model of the urban energy system under disaster events is constructed. Based on the preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under disaster events is converted into a mixed integer linear constraint, and the mixed integer linear constraint is embedded in the two-layer optimization model of the urban energy system to obtain the optimization operation simulation result under the mixed integer linear constraint. Therefore, in order to solve the problem that related technologies, on the one hand, ignore the emergency supply capacity and buffering effect of hot / cold pipelines in disaster events, and on the other hand, lead to large-scale optimization problems and high solution complexity, a rule with simple form, strong interpretability, strong operability and safety is provided, which can significantly improve the solution efficiency and has significant innovation and good application value.

[0069] Specifically, Figure 1 A flow chart of a method for extracting and optimizing security rules for improving the resilience of urban energy systems provided in an embodiment of the present application.

[0070] like Figure 1 As shown, the security rule extraction and optimization method for improving the resilience of urban energy systems includes the following steps:

[0071] In step S101, the resilience index of the urban energy system is calculated, and a two-layer optimization model of the urban energy system is constructed based on the resilience index.

[0072] Specifically, the resilience of an urban energy system is the ability to ensure the continuous supply of all critical loads in the event of a disaster. After an extreme disaster occurs, the supply of some non-critical loads should be temporarily abandoned as appropriate, giving priority to ensuring the continuous supply of critical loads. Therefore, the actual supply load of the system is used to characterize the system performance, and the maximum degree of decline in system performance is used to calculate the resilience index. The expression is as follows:

[0073]

[0074] Among them, SP0 is the system performance before the disaster event occurs, SP d is the minimum value of system performance after a disaster event, P load It is the system load (including important load and non-important load).

[0075] When the actual load supplied by the system after a disaster event is always higher than the sum of all important loads, the system is considered to have resilience to cope with the disaster event, otherwise it is not resilient. The resilience constraint is:

[0076]

[0077] in, It is an important load of the system.

[0078] Based on the above definition of resilience indicators, a two-level optimization model for the urban energy system is constructed. The upper level problem (3) is the problem of normal system operation, seeking the system operation instructions with the best economic efficiency; the lower level problem (4) is the problem of system resilience operation under disaster events, seeking the system operation instructions with the least load shedding, while also meeting the resilience constraints, that is, ensuring the continuous supply of important loads.

[0079]

[0080] stg(x(t))≤0,

[0081]

[0082]

[0083] stl(x(t′),x(t0),y)≤0,

[0084]

[0085]

[0086] Where f(·), h(·) and g(·)≤0, l(·)≤0 are the objective functions and operation constraints of the upper and lower problems, respectively; x is the system operation state variable; y is the system damage variable caused by the disaster event; t and t′ are time parameters; T is the duration of the upper problem scenario; t0 is the time when the disaster event occurs; T′ is the system repair time after the disaster event; Ω y It is a collection of disaster event scenes.

[0087] In fact, after a disaster occurs, the system only cares about whether the important loads can be continuously supplied. Therefore, the main purpose of the lower-level problem is not to solve the optimal operation plan, but to solve the feasible operation plan that meets the resilience constraints. Therefore, the lower-level problem is essentially a reflection of the system operation status in the upper-level problem. To this end, a data-driven method is proposed to fit the constraints of the lower layer on the upper layer with a series of mixed integer linear constraints, replacing the original lower layer problem:

[0088]

[0089]

[0090] Among them, A i The coefficient representing the operating state x(t) of device i; b i I represents the safe operation boundary of device i obtained by data driving; i is a binary variable, representing the fault status of device i, 1 means running, 0 means fault; M is a positive large number. Therefore, the original two-level optimization problem (3)-(4) can be transformed into:

[0091]

[0092] stg(x(t))≤0,

[0093] A i x(t)≤b i +M·(1-I i ),i=1,2,...,G

[0094]

[0095]

[0096] In step S102, based on the preset energy transmission link, the preset energy storage link and the preset energy conversion link, the dynamic characteristics of the power electronics link of the urban energy system are modeled to obtain a dynamic characteristics model of the power electronics link.

[0097] Specifically, in the energy transmission link: Flexible soft switching technology replaces traditional feeder-based circuit breaker-based feeder interconnectors with controllable power electronic converters, achieving normalized flexible soft connections between feeders. After a system fault occurs, the flexible soft switch responds quickly, enabling rapid network reconstruction and load transfer of the distribution network, prioritizing the power supply of important loads. The flexible soft switch can simultaneously operate when a system fault occurs:

[0098] P SOP (t)≤M·(1-S line (t)); (7)

[0099] 0≤P SOP (t)≤P SOP,max ; (8)

[0100] Among them, P SOP and P SOP,max They are the operating power and upper limit of the flexible soft switch; S line It is the operating status of the distribution line, which is a 0-1 variable, 1 indicates that the line is operating normally, and 0 indicates that the line is faulty; M is a positive number.

[0101] Energy storage: Batteries are connected to the grid through power electronic converters, and the battery's charge and discharge power can be rapidly adjusted by the power electronic converters. After a system failure occurs, the battery can quickly increase its discharge power, effectively connecting to a backup power source to support critical loads before the power is fully discharged. This provides an effective buffer for the operation of other equipment in the urban energy system and ensures the continuous supply of critical loads after a failure. The battery discharge power can be instantly adjusted at the moment of a system failure:

[0102] 0≤P ES (t0)≤P ES,max ; (9)

[0103] Among them, P ES and P ES,max They are the operating power of the battery and its upper limit.

[0104] Energy conversion: Heat pumps and compression chillers utilize a small amount of electrical energy to drive a compressor, collecting large amounts of heat from the environment. These devices are electricity-to-heat and electricity-to-cooling energy conversion devices. After a power system failure, due to the thermal inertia of the heating / cooling system, heat pumps and compression chillers can rapidly reduce their operating power through frequency conversion, achieving rapid demand response and prioritizing the supply of critical loads in the power system. The input power of heat pumps and compression chillers can be instantly adjusted at the moment of a system failure:

[0105] 0≤P EHP (t0)≤P EHP,max ; (10)

[0106] 0≤P CERG (t0)≤P CERG,max ; (11)

[0107] Among them, P EHP 、P CERG and P EHP,max 、P CERG,max They are the operating power and upper limits of heat pumps and compression chillers respectively.

[0108] In addition, the power electronics link also has the ability to operate in a short-term overload state. In theory, the shorter the overload operation time of the power electronics, the stronger its overload capacity. The flexible soft switch and battery meet the following requirements during overload operation:

[0109] 0≤P SOP (t)≤P SOP,max α SOP,over ; (12)

[0110] 0≤P ES (t)≤P ES,max α ES,over ; (13)

[0111] Among them, α SOP,over and α ES,over Characterize the overload capacity of the flexible soft switch and the battery respectively.

[0112] In step S103, based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas grid and the storage capacity of the district heating / cooling network pipelines, the multi-energy network dynamic characteristics of the urban energy system are modeled to obtain a multi-energy network dynamic characteristics model.

[0113] Optionally, in some embodiments, based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas grid and the storage capacity of the regional heating / cooling network pipelines, the dynamic characteristics of the multi-energy network of the urban energy system are modeled to obtain a multi-energy network dynamic characteristics model, including: based on the storage capacity of the natural gas grid and the storage capacity of the regional heating / cooling network pipelines, determining a first relationship between the energy reserves of the pipeline network equivalent energy storage at the time of the disaster event and the operating status of the heat source at the time of the disaster event; based on the real-time balance of supply and demand of the power system, determining a second relationship between the energy reserves of the load thermal inertia equivalent energy storage at the time of the disaster event and the heat / cooling load at the time of the disaster event; based on the first relationship and the second relationship, the dynamic characteristics of the multi-energy network of the urban energy system are modeled to obtain a multi-energy network dynamic characteristics model.

[0114] Specifically, unlike distribution networks, natural gas networks and district heating / cooling network pipelines have a certain storage capacity, and the amount of stored energy can be changed by changing the gas pressure and working fluid temperature respectively. Taking the heating network as an example, when the heat source fails and the downstream heating pipeline network is still operating normally, the heat stored in the working fluid in the heating network pipeline will not disappear immediately. At this time, the heating pipeline network can be equivalent to a heat storage device that begins to release the stored energy until the working fluid temperature drops to the minimum limit. The mathematical model of the equivalent energy storage of the pipeline network is:

[0115] E l ·(SOC l (t+1)-SOC l (t))=-P l (t) / η l ·Δt; (14)

[0116] Among them, E l , SOC l 、P l and η l They are the upper limit of the equivalent energy storage of the pipeline network, the state of charge (SOC), the release power and the efficiency.

[0117] The equivalent energy storage of the pipeline network is related to the average temperature of the pipeline working fluid at that moment:

[0118] E l =c p m l ·(T l (t0)-T l,min ); (15)

[0119] Among them, c p is the specific heat capacity of the working fluid; m l is the quality of the pipeline working fluid; T l is the average temperature of the working fluid in the pipeline; T l,minIt is the minimum limit of pipe temperature.

[0120] On the other hand, the working fluid temperature in the heating network (including the water supply and return pipes) is related to the heating power of the heat source. Taking a heating network with a gas boiler as the sole heat source as an example, there is a relationship between the thermal power input into the heating network by the gas boiler and the working fluid temperature injected into the water supply pipe at the gas boiler and the working fluid temperature flowing out of the return pipe:

[0121] P AB (t) = c p v m (T l,AB (t)-T lr,AB (t)); (16)

[0122] Among them, P AB is the operating power of the gas boiler; T l,AB and T lr,AB are the working medium temperature injected into the water supply pipe at the gas boiler and the working medium temperature flowing out of the return pipe; v m is the mass flow rate of the working fluid.

[0123] If the transmission heat loss of the heating network is ignored, the thermal power input to the heating network by the gas boiler can be calculated using the average temperature difference of the working medium in the supply pipe and the return pipe:

[0124] P AB (t) = c p v m (T l (t)-T lr (t)); (17)

[0125] Among them, T lr is the average temperature of the working fluid in the return water network.

[0126] From the combined equations (15) and (17), we can see that the energy storage E of the equivalent energy storage of the pipeline network at the time of the disaster event is l The operating state P of the heat source at that moment AB (t0) is associated.

[0127] Power systems require a real-time balance between supply and demand. However, some heating / cooling loads (such as building space heating / cooling loads) exhibit thermal inertia with respect to changes in heating / cooling supply. This means that when a heating / cooling source fails, these heating / cooling loads do not immediately fail. The time difference between the failure of a heating / cooling source and the failure of a heating / cooling load provides an effective buffer for urban energy systems during disaster events. The building space heating model is:

[0128] Q(t+1)-Q(t)=C b ·(T b (t+1)-T b (t))=(P in(t)-P load (t))·Δt; (18)

[0129] Where Q is the heat stored in the building's indoor space; C b is the heat capacity of the building; T b is the indoor temperature of the building; P in is the thermal power provided by the heat source.

[0130] During normal operation, the heating power is equal to the heat load and the indoor temperature remains unchanged. When the heating system fails, the heating power is zero, then:

[0131] P load (t)·Δt=P disch (t)·Δt=Q(t)-Q(t+1)=C b ·(T b (t)-T b (t+1)); (19)

[0132] Among them, P disch is the release power of the building's heat storage.

[0133] According to Equation (19), the thermal inertia of the heating / cooling load can be modeled as an equivalent heat / cooling device, which starts to release energy when the heating / cooling load cannot be supplied normally:

[0134] E b ·(SOC b (t+1)-SOC b (t))=-P disch (t)·Δt; (20)

[0135] Among them, E b and SOC b They are the upper limit of storage and storage status of thermal inertia equivalent energy storage of hot / cold load respectively.

[0136] When the indoor temperature leaves the comfortable temperature range, the heating / cooling load is considered to be ineffective. Therefore, the energy storage of the load thermal inertia equivalent energy storage at the time of the disaster event only models the temperature changes within the comfortable temperature range:

[0137] E b =C b ·|T b (t0)-T limit |; (21)

[0138] Among them, T limit It is the comfort temperature critical value (the lower limit value is used for heating load and the upper limit value is used for cooling load).

[0139] On the other hand, the size of the heating / cooling load (i.e. the heating / cooling demand of the building) is related to the temperature difference between the indoor and outdoor areas of the building:

[0140] P load (t) = K b ·|T b (t)-T a |; (22)

[0141] Among them, K b is the building heat dissipation coefficient; T a is the outdoor ambient temperature.

[0142] From the combined equations (21) and (22), we can see that the energy storage E of the load thermal inertia equivalent energy storage at the time of the disaster event is b The heating / cooling load P at that moment load (t0)(or the building indoor temperature setting T b (t0)) is associated.

[0143] In step S104, based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, a two-stage resilience operation model of the urban energy system under disaster events is constructed, and based on the preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under disaster events is converted into a mixed integer linear constraint, and the mixed integer linear constraint is embedded in the two-layer optimization model of the urban energy system to obtain the optimization operation simulation results under the mixed integer linear constraint.

[0144] According to the longest overload time of power electronic equipment, the disaster event is divided into two stages:

[0145] Phase 1: The immediate period after the disaster. This phase spans [t0, t1] and typically occurs on a minute-scale. During this phase, the fast dynamic characteristics of power electronics play a dominant role. Power electronics rapidly respond and briefly overload to support the supply of critical system loads. During this period, the rest of the urban energy system is unable to react quickly in the short period following the disaster.

[0146] Phase 2: The long period before the system is fully repaired. This phase spans [t1, t2], typically on an hourly scale, and typically lasts from a few hours to a day, or even several days. During this phase, the slow dynamic characteristics of the multi-energy network play a dominant role. Thermal storage / cooling equipment, equivalent energy storage for heating / cooling pipe networks, and equivalent energy storage for thermal inertia of heat / cooling loads provide emergency supply during system repair. Power electronics have lost their overload capability during this phase.

[0147] The objective function of the two-stage resilience operation model of the urban energy system is to minimize the total amount of system load reduction under disaster events:

[0148]

[0149] Among them, LS k is the load loss magnitude of load k; Ω L is the load set.

[0150] Phase 1: Shortly after the disaster

[0151] The constraints at this stage include: power conservation constraints, energy conversion equipment operation constraints, energy storage equipment operation constraints, energy transmission equipment operation constraints, equipment capacity constraints, system input constraints, power electronics link fast dynamic characteristics constraints, equipment operation status continuity constraints when disaster events occur, equipment failure constraints, and system resilience constraints.

[0152] Power conservation constraint. For any device, its input power equals the sum of the output powers at the end of all lines connected to its input, and its output power equals the sum of the input powers at the beginning of all lines connected to its output. For any bus, the sum of its input power equals the sum of its output power.

[0153]

[0154]

[0155]

[0156] in, and are the input power and output power of device i, respectively; and are the input power at the beginning and output power at the end of line j respectively; and are the sets of lines connected to the input and output of device i respectively; and are the sets of lines feeding into and out of busbar b; Ω u and Ω b They are respectively a collection of unit equipment and busbars.

[0157] Energy conversion equipment operating constraints. Use energy conversion efficiency to model the relationship between input power and output power of energy conversion equipment:

[0158]

[0159] Among them, η i is the energy conversion efficiency of energy conversion device i; Ω uc It is a collection of energy conversion devices.

[0160] For single-input multi-output energy conversion equipment, such as combined heat and power units, multiple energy conversion efficiencies are introduced:

[0161]

[0162]

[0163] Among them, η g2p and η g2h They are gas-to-electricity conversion efficiency and gas-to-heat conversion efficiency respectively.

[0164] Energy storage device operation constraints. In addition to the charge and discharge power, the storage state is introduced to describe the operation of the energy storage device:

[0165]

[0166] Among them, E i , SOC i 、 and They are the storage limit, storage status, charging power, discharging power, charging efficiency and discharging efficiency of the energy storage device; Ω us It is a collection of energy storage devices.

[0167] Energy transmission equipment operation constraints. Use operational efficiency to model the energy transmission losses of energy transmission equipment (including distribution lines, natural gas pipelines, and heating / cooling pipelines):

[0168]

[0169] in, is the operating efficiency of line j; Ω l is a collection of lines.

[0170] Equipment capacity constraints. The operating status of each device in the system (including energy conversion equipment, energy storage equipment, new energy power generation equipment, and energy transmission equipment) is subject to its construction capacity constraints:

[0171]

[0172]

[0173]

[0174] Among them, C i is the capacity of device i; Ω c Is a collection of system devices; Ω ur It is a collection of new energy power generation equipment; PU i is the per-unit output curve of new energy power generation equipment i.

[0175] System input constraints. The energy input from outside the region is constrained by the capacity of the tie line:

[0176]

[0177] in, and are the energy of input energy k per unit time and its upper limit; Ω buy is the set of input energies.

[0178] Constraints on the fast dynamic characteristics of the power electronics link. Within a short period of time after a disaster event, the operating state of the power electronics link can change rapidly and may even operate overloaded. See Equations (7) to (13).

[0179] Continuity constraints on equipment operating status during a disaster event. Unlike flexible switches, batteries, heat pumps, and compression chillers, the operating status of other equipment (including cogeneration units, gas boilers, absorption chillers, thermal storage equipment, and cold storage equipment) will not change suddenly before or after a disaster event. Taking a cogeneration unit as an example:

[0180]

[0181] In addition, the reserves of various energy storages in the system (including batteries, heat / cold storage equipment, heating / cold pipe network equivalent energy storage, heat / cold load thermal inertia equivalent energy storage, etc.) will not change suddenly before and after the disaster event. Taking batteries as an example:

[0182] SOC ES (t 0+ )=SOC ES (t 0- );(37)

[0183] Equipment failure constraints. The operating point of each device in the system is also constrained by its fault state (operating / faulty):

[0184]

[0185] Among them, S i is the fault status of device i, 1 indicates operation and 0 indicates fault; M is a positive large number.

[0186] System resilience constraint. The system resilience index must be less than the set limit to ensure the continuous supply of important loads. See formula (2).

[0187] Phase 2: The long period before the system is fully repaired

[0188] This stage also needs to meet the following constraints: power conservation constraints, energy conversion equipment operation constraints, energy storage equipment operation constraints, energy transmission equipment operation constraints, equipment capacity constraints, system input constraints, equipment failure constraints and system resilience constraints.

[0189] At this stage, the power electronic link (including the flexible soft switch and battery) has lost its overload operation capability, and its operation only needs to meet the equipment capacity constraint, see formula (32).

[0190] In addition, the slow dynamic characteristics constraints of the multi-energy network need to be considered, including the equivalent energy storage constraints of the pipeline network and the equivalent energy storage constraints of the thermal inertia of hot / cold loads.

[0191] Pipeline network equivalent energy storage constraint. If a heat / cooling source fails during a disaster and the downstream heating / cooling network still operates normally, the downstream heating / cooling network can be equivalent to a heat / cooling storage device. See Equations (14)-(17).

[0192] Equivalent energy storage constraint for thermal inertia of hot / cold loads. Use equivalent heat / cold storage devices to model the thermal inertia of hot / cold loads. See Equations (20) to (22).

[0193] Optionally, in some embodiments, based on a preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under disaster events is converted into mixed integer linear constraints, including: obtaining a sparse weight oblique decision tree based on the preset sparse weight oblique decision tree model; using the sparse weight oblique decision tree to optimize the objective function of each node in the decision tree to convert the two-stage resilience model of the urban energy system under disaster events into mixed integer linear constraints.

[0194] Specifically, a decision tree is a supervised learning model that uses a tree structure to express any functional relationship between learning objectives and features. Based on the univariate decision tree, the oblique decision tree (also known as the multivariate decision tree) uses a hyperplane to linearly divide the nodes of the decision tree into θ T p replaces the original single variable division, effectively improving the learning performance of the decision tree. T p<0, then the running state p belongs to the left child node; if θ T p≥0, then the running state p belongs to the right child node. The oblique decision tree also takes minimizing the impurity index as the optimization goal to obtain the optimal partition parameter vector θ, but due to the indicative function I[θ T p<0] is discontinuous, which makes obtaining the optimal partitioning parameters a discrete optimization problem. To this end, some scholars proposed a weighted oblique decision tree to solve the optimal partitioning parameters with weighted information entropy as the objective function. The core idea is to use the sigmoid function σ(θ T The "soft partitioning" of p) replaces the characteristic function l[θ TThe "hard partition" of p<0] is that any sample enters the left child node with a certain probability (i.e., the weight of the left child node) and enters the right child node with the remaining probability (i.e., the weight of the right child node), achieving continuous and smooth impurity index function, thereby transforming the discrete optimization problem into a continuous optimization problem. The expression of weighted information entropy E(θ) is:

[0195] E(θ)=W L (θ)H L (θ)+W R (θ)H R (θ); (39)

[0196] Among them, W L (θ) and W R (θ) is the sum of the weights of the left and right child nodes, respectively, H L (θ) and H R (θ) are the weighted information entropy of the left child node and the right child node respectively.

[0197] However, the limitation of the weighted oblique decision tree is that the optimal partitioning parameters it obtains are dense vectors, and the resulting security rules are not easy to understand. To this end, some scholars have further proposed a sparse weighted oblique decision tree, using weighted information entropy and ElasticNet regularization terms as the optimization objective function to solve the optimal partitioning parameters. The core idea is that the Elastic Net regularization term can set the feature quantity coefficients that are not important to system security to zero, while ensuring the correlation between important feature quantity coefficients, making the security rules easy to understand. The optimization objective function of the sparse weighted oblique decision tree at each node is:

[0198]

[0199] Among them, the first term is the weighted information entropy of the current node, which is normalized by the number of samples N of the current node; the sum of the second and third terms is the Elastic Net regularization term. This optimization problem can be solved by the improved quadrant quasi-Newton method.

[0200] Then, security rule extraction and embedded optimization are performed. The premise of security rule extraction is to analyze the feature quantities contained in the system operation status p, which mainly consists of the following five types of feature quantities:

[0201] 1. Output p of energy conversion and energy storage equipment. On the one hand, the output of equipment such as cogeneration units, gas boilers, absorption chillers, thermal storage equipment, and cold storage equipment does not change suddenly before or after a disaster event, and is directly related to the system's energy supply in the immediate aftermath. On the other hand, as shown in Section 2.3, the energy reserves of the equivalent energy storage of the heating / cooling pipeline network at the time of a disaster are related to the output of its upstream heat / cooling source at that moment. The output levels of energy conversion and energy storage equipment also reflect the reserves of the equivalent energy storage of the pipeline network.

[0202] 2. Energy storage capacity (SOC). The capacity of batteries, thermal storage devices, and cold storage devices at the time of a disaster directly determines their ability to provide emergency supplies to the system and how long they can last.

[0203] 3. Building Indoor Temperature TEMP. As can be seen from the energy conversion process in step S102, the energy reserve of the thermal inertia equivalent energy storage of the heating / cooling load at the time of the disaster event is related to the building indoor temperature setting at that moment. This equivalent energy storage reflects the margin for the heating / cooling load to be temporarily disconnected from the heat / cooling source.

[0204] 4. Repair time T of each device in the system re Equipment repair time is determined by factors such as the extent of damage caused by the disaster, the type of equipment, and repair efficiency. Equipment repair time can directly affect the system's emergency supply capabilities.

[0205] 5. Critical load level D. The critical load level at the time of a disaster event can directly affect the short-term support and buffering capacity of the post-disaster system.

[0206] In summary, the system operating state p can be written as:

[0207] p=[P T ,SOC T ,TEMP T ,T re T ,D T ,1] T ; (41)

[0208] Among them, the last feature is always 1 and is used to learn the bias term of the linear hyperplane.

[0209] Let safety label y∈{b,g}, where b represents unsafe operation state and g represents safe operation state. Then the normal operation state dataset of the system with safety label y is:

[0210] Ω p ={(p1,y1),(p2,y2),...,(p N ,y N )}; (42)

[0211] The system normal operating status dataset with safety label y is input into the sparse weight oblique decision tree model to train and generate a decision tree.

[0212] According to the decision tree structure, any system operating state p has only one corresponding leaf node. Each leaf node classified as a safe operating state represents a safety rule, and the union of the safety rules represented by these leaf nodes is the complete safety rule. The safety rules represented by any two leaf nodes are mutually exclusive. In order to extract the complete safety rules, it is necessary to find all leaf nodes classified as safe operating states and obtain the safety rules represented by these leaf nodes. Based on the recursive idea, Algorithm 1 gives a specific method for extracting safety rules. Assume that the decision tree has a total of G leaf nodes classified as safe operating states, and R i is the rule matrix of the i-th leaf node classified as safe operation state, then the extracted safety rules can be expressed as:

[0213] R i p≥0,i=1,2,...,G; (43)

[0214] Table 1

[0215]

[0216]

[0217] Embedded optimization operation. Since the safety rules represented by any two leaf nodes are mutually exclusive, it is impossible to directly embed the safety rules shown in Equation (43) as constraints into the normal operation model of the urban energy system. This paper adopts the Big M method to introduce auxiliary 0-1 variables to transform the safety rules into constraints:

[0218] R1p≥-M(1-I1)

[0219] R2p≥-M(1-I2) ...

[0221] R G p≥-M(1-I G )

[0222]

[0223] Among them, I i is a 0-1 variable corresponding to the leaf node classified as safe operation state. i= 1, then the system operating state p satisfies the safety rule corresponding to the i-th leaf node classified as a safe operating state; otherwise, it does not. M is a large positive number. The last line of the constraint in Equation (44) indicates that the system operating state p satisfies only one of the safety rules, and this operating state is a safe operating state.

[0224] It should be noted that the repair time T of each device in the system re Characteristic quantities such as the important load level D are difficult to be used as real-time adjustment objects, so these characteristic quantities are set as established parameters constrained by safety rules.

[0225] By embedding Equation (44) as a constraint into the normal operation model of the urban energy system, we can obtain optimized operation simulation results under the constraints of safety rules, taking into account both the economic and safety aspects of the urban energy system operation. Equation (44) is a mixed-integer linear constraint on the decision variables. Therefore, the normal operation model with embedded safety rules can be transformed into a standard mixed-integer linear programming problem, which can be efficiently solved using commercial solvers (such as Cplex and Gurobi).

[0226] In order to enable those skilled in the art to further understand the security rule extraction and optimization method for improving the resilience of urban energy systems according to the embodiments of the present application, it is elaborated in detail below with reference to specific embodiments.

[0227] like Figure 2 As shown, Figure 2 This is a flowchart of a security rule extraction and optimization method for improving the resilience of urban energy systems proposed in an embodiment of the present application, which includes the following five main steps.

[0228] Step 1: Simulate normal operation. Perform a system operation simulation under normal scenarios and record a large dataset of normal system operation status. Each data point corresponds to a simulation result of a normal operation scenario.

[0229] Step 2: Generate a safety label for the dataset. Using a data point from the system's normal operating state dataset as a boundary condition, solve the system's resilience model under a disaster event. If the problem has a feasible solution, the data point is labeled "safe," otherwise, it is labeled "unsafe." This operation is repeated across the system's normal operating state dataset to generate a safety-labeled (safe / unsafe) normal operating state dataset.

[0230] Step 3: Extract safety rules. Using the normal operating status dataset with safety labels as input, we perform binary classification learning with "safe" and "unsafe" as the two categories. After learning, we output the classification boundary, where the boundary of the "safe" category is the safety rule in Equation (5).

[0231] Step 4: Embed security rules. Embed security rules into the system's normal operation model, solve the system's normal operation model with security rule constraints, and then perform a large number of normal scenario simulations again to record a large data set of the system's normal operation status considering security rules.

[0232] Step 5: Verify the effectiveness of safety rules. Using a data point from the system's normal operating state dataset with safety rules as a boundary condition, solve the system resilience operation model under a disaster event separately and calculate the probability that the problem has a feasible solution at this time.

[0233] The system of the embodiment of the present application consists of two areas (hereinafter referred to as area 1 and area 2), such as Figure 3 As shown in the figure, both areas have their own power and natural gas sources on the supply side; their own energy conversion equipment, energy storage equipment, energy transmission equipment, renewable energy power generation equipment, substations, heat exchange stations, etc. on the network side; and their own electricity, heating, and cooling loads on the demand side. The distribution network adopts a single-ring structure, with a tie switch installed between Areas 1 and 2. During normal system operation, the tie switch is open, and Areas 1 and 2 are self-sufficient in load supply. In the event of a system failure, the electrical load can be transferred by closing the tie switch.

[0234] The electricity, heating and cooling load curves of the two areas are as follows: Figure 4 As shown in the figure, the heating and cooling loads in this load curve exhibit distinct seasonal characteristics, with heating loads primarily concentrated in winter and cooling loads primarily concentrated in summer. The critical loads for electricity, heating, and cooling account for 50%, 40%, and 20% of the total load, respectively. The upper limit for power and natural gas supplied by the main grid to Areas 1 and 2 is 200MWh per hour. The main grid electricity price adopts a time-of-use (TOU) pricing system, with peak, flat, and off-peak hour prices of 1,322 yuan / MWh, 839 yuan / MWh, and 381 yuan / MWh, respectively. The peak summer price is 1,440 yuan / MWh. Natural gas is priced at 300 yuan / MWh.

[0235] Table 2 shows the construction capacity and operating efficiency of various energy conversion equipment, energy storage equipment and new energy power generation equipment. Figure 5 The annual power generation data of the photovoltaic power generation system is given.

[0236] Table 2 Equipment construction capacity and operation efficiency

[0237]

[0238]

[0239] Taking Area 1 as an example, we extract the system operation safety rules for the summer (June, July, and August), assuming that the energy storage equipment and the heat / cold network energy storage are available after the disaster event. All eigenvalues ​​are normalized before entering the sparse weight oblique decision tree model. The safe and unsafe operation states are represented by binary variables. In addition, the longest repair time of the nodes and branches in series between the two busbars is input into the decision tree model. Only one characteristic value T is input for each series route. re .

[0240] Figure 6 The system operation safety rules are shown (depth 3, regularization coefficient 0.001). The subscripts CHP, AB, EHP, CERG, WARG, ES, HS, CS, HL, CL, RES, and line correspond to cogeneration unit, gas boiler, heat pump, compression chiller, absorption chiller, battery, thermal storage equipment, cold storage equipment, heating load, cooling load, renewable energy power generation equipment, and distribution line, respectively.

[0241] 200 system operation states are randomly generated in winter and summer. The calculation results of the safe operation state ratio of the three cases of normal operation, embedded safety rules, and two-layer optimization model are as follows: Figure 7 As shown in Figure 2 , when disaster response is not considered, the system's safe operating state ratio under disaster events does not exceed 70%. After embedding safety rules, the system's safe operating state ratio increases to over 96%, verifying the effectiveness of the proposed method. The two-level optimization model can achieve a 100% safe operating state ratio, which is equivalent to embedding a safety verification mechanism model. Therefore, the data-driven approach better approximates the performance of the model-driven approach with fewer mixed-integer linear constraints.

[0242] Figure 8 The system optimization operation cost (i.e., objective function value) and calculation time of typical days in winter and summer (typical days are obtained by k-center value clustering method) under three scenarios are shown. When disaster response events are not considered, the system operation cost is the lowest. By embedding safety rules ( Figure 8 (a)) and the two-level optimization model ( Figure 8After the two methods (b) constrained the system operating state, the system operating cost was improved, and the error of the two methods did not exceed 3%. In terms of computing time, the method of embedding safety rules only introduced a small number of mixed integer linear constraints on the basis of the normal operation model of the system, and its computing time was almost unchanged, both at 0.6 to 0.7 seconds. However, the two-layer optimization model was solved using a stochastic optimization method, and the computing time exceeded 180 seconds when the lower layer problem considered 50 disaster event scenarios. Therefore, by extracting and embedding safety rule constraints through a data-driven method, the computing time can be reduced by 2 orders of magnitude, when the error between the optimization results and the model-driven method is less than 3%, and it is basically the same as the solution time of the normal operation problem.

[0243] Therefore, the embodiment of the present application proposes a two-stage resilience operation model for the urban energy system under disaster events, which improves the resilience of the urban energy system under disaster events by coordinating the resilience resources of the two time scales of the fast dynamic characteristics of power electronic equipment and the slow dynamic characteristics of the multi-energy network. Phase one (the short-term stage after the disaster event) mainly considers the short-term support role of the rapid action and short-term overload capacity of power electronic equipment on important loads; Phase two (the long-term stage before the system is fully repaired) mainly considers the buffering effect of the pipeline energy storage and load thermal inertia of the multi-energy network under disaster events. In addition, a data-driven urban energy system safety rule extraction and embedded optimization operation method is proposed. Based on the sparse weight oblique decision tree theory, the two-stage resilience model of the urban energy system under disaster events is converted into a small number of mixed integer linear constraints, namely safety rules, and embedded into the normal operation model of the system. The extracted safety rules are simple in form, highly interpretable, and highly operational, and can significantly improve the solution efficiency, with significant innovation and good application value.

[0244] According to the safety rule extraction and optimization method for improving the resilience of urban energy systems proposed in the embodiment of the present application, by calculating the resilience index of the urban energy system and constructing a two-layer optimization model of the urban energy system based on the resilience index, the dynamic characteristics of the power electronics link of the urban energy system are modeled based on the preset energy transmission link, the preset energy storage link and the preset energy conversion link to obtain the dynamic characteristics model of the power electronics link, and based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipelines, the dynamic characteristics of the multi-energy network of the urban energy system are modeled to obtain the dynamic characteristics model of the multi-energy network, and based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, a two-stage resilience operation model of the urban energy system under disaster events is constructed, and based on the preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under disaster events is converted into a mixed integer linear constraint, and the mixed integer linear constraint is embedded in the two-layer optimization model of the urban energy system to obtain the optimization operation simulation result under the mixed integer linear constraint. This solves the problem that related technologies, on the one hand, ignore the emergency supply capacity and buffering effect of hot / cold pipelines in disaster events, and on the other hand, lead to large-scale optimization problems and high solution complexity. It provides rules with simple form, strong interpretability, strong operability and safety, which can significantly improve the solution efficiency and have significant innovation and good application value.

[0245] Next, a security rule extraction and optimization device for improving the resilience of urban energy systems proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0246] Figure 9 It is a block diagram of a security rule extraction and optimization device for improving the resilience of urban energy systems according to an embodiment of the present application.

[0247] like Figure 9 As shown, the security rule extraction and optimization device 10 for improving the resilience of urban energy systems includes: a calculation module 100, a generation module 200, a modeling module 300 and an optimization module 400.

[0248] Among them, the calculation module 100 is used to calculate the resilience index of the urban energy system and construct a two-layer optimization model of the urban energy system based on the resilience index; the generation module 200 models the dynamic characteristics of the power electronics link of the urban energy system based on the preset energy transmission link, the preset energy storage link and the preset energy conversion link, and obtains the dynamic characteristics model of the power electronics link; the modeling module 300 is used to model the dynamic characteristics of the multi-energy network of the urban energy system based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipeline, and obtains the dynamic characteristics model of the multi-energy network; the optimization module 400 is used to construct a two-stage resilience operation model of the urban energy system under disaster events based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, and based on the preset sparse weight oblique decision tree model, convert the two-stage resilience model of the urban energy system under disaster events into a mixed integer linear constraint, and embed the mixed integer linear constraint into the two-layer optimization model of the urban energy system to obtain the optimization operation simulation result under the mixed integer linear constraint.

[0249] Optionally, in some embodiments, the calculation module 100 is further configured to calculate the urban energy system resilience index based on a preset resilience index formula, wherein the preset resilience index formula is:

[0250]

[0251] Among them, ε R is the resilience index, SP0 is the system performance before the disaster event, SP d is the minimum value of system performance after a disaster event, P load is the system load.

[0252] Optionally, in some embodiments, the generating module 200 is further configured to determine a preset energy transmission link based on a first instantaneous adjustment strategy of a response speed of the flexible soft switch, wherein the first instantaneous adjustment strategy is:

[0253] P SOP (t)≤M·(1-S line (t));

[0254] 0≤P SOP (t)≤P SOP,max ;

[0255] in, is the important load of the system, P SOP () is the operating power of the flexible soft switch at time t, M is a positive number, S line is the operating status of the distribution line, P SOP,max The upper limit of the operating power of the flexible soft switch;

[0256] Based on the second instantaneous regulation strategy of the discharge power of the battery, a preset energy storage link is determined, wherein the second instantaneous regulation strategy is:

[0257] 0≤P ES (t0)≤P ES,max ;

[0258] Among them, P ES is the operating power of the battery, P ES, is the upper limit of the battery's operating power;

[0259] Based on the third instantaneous regulation strategy of the input electric power of the heat pump and the input electric power of the compression refrigerator, a preset energy conversion link is determined, wherein the third instantaneous regulation strategy is:

[0260] 0≤P EHP (t0)≤P EHP,max ;

[0261] 0≤P CERG (t0)≤P CERG,max ;

[0262] Among them, P EHP is the heat pump operating power, P EHP,max is the upper limit of the heat pump operating power, P CERG is the operating power of the compression refrigerator, P CERG, It is the upper limit of the operating power of the compression refrigerator.

[0263] Optionally, in some embodiments, the modeling module 300 is further used to: determine a first relationship between the energy reserves of the pipeline network equivalent energy storage at the time of the disaster event and the operating status of the heat source at the time of the disaster event based on the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipeline; determine a second relationship between the energy reserves of the load thermal inertia equivalent energy storage at the time of the disaster event and the heat / cooling load at the time of the disaster event based on the real-time balance of supply and demand of the power system; model the multi-energy network dynamic characteristics of the urban energy system according to the first relationship and the second relationship to obtain a multi-energy network dynamic characteristics model.

[0264] Optionally, in some embodiments, the objective function of the two-stage resilience operation model of the urban energy system is:

[0265]

[0266] Where k is the load, LS k is the load loss size of load k, Ω L is the load set, and t is the time.

[0267] Optionally, in some embodiments, the optimization module 400 is also used to: obtain a sparse weight oblique decision tree based on a preset sparse weight oblique decision tree model; and use the optimization objective function of each node of the sparse weight oblique decision tree to convert the two-stage resilience model of the urban energy system under disaster events into a mixed integer linear constraint.

[0268] Optionally, in some embodiments, the optimization objective function of the sparse weight oblique decision tree at each node is:

[0269]

[0270] Among them, θ is the optimal partition parameter vector, N is the number of samples, W L () is the weight sum of the left child node, W R () is the weight sum of the right child node, H L () is the weighted information entropy of the left child node, H R () is the weighted information entropy of the right child node, λ1|θ|+2‖θ‖ 2 is the Elastic Net regularization term.

[0271] It should be noted that the above explanation of the embodiment of the security rule extraction and optimization method for improving the resilience of urban energy systems is also applicable to the security rule extraction and optimization device for improving the resilience of urban energy systems in this embodiment, and will not be repeated here.

[0272] According to the safety rule extraction and optimization device for improving the resilience of urban energy systems proposed in the embodiment of the present application, by calculating the resilience index of the urban energy system and constructing a two-layer optimization model of the urban energy system based on the resilience index, the dynamic characteristics of the power electronics link of the urban energy system are modeled based on the preset energy transmission link, the preset energy storage link and the preset energy conversion link to obtain the dynamic characteristics model of the power electronics link, and based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network and the storage capacity of the regional heating / cooling network pipelines, the dynamic characteristics of the multi-energy network of the urban energy system are modeled to obtain the dynamic characteristics model of the multi-energy network, and based on the dynamic characteristics model of the power electronics link and the dynamic characteristics model of the multi-energy network, a two-stage resilience operation model of the urban energy system under disaster events is constructed, and based on the preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under disaster events is converted into a mixed integer linear constraint, and the mixed integer linear constraint is embedded in the two-layer optimization model of the urban energy system to obtain the optimization operation simulation result under the mixed integer linear constraint. This solves the problem that related technologies, on the one hand, ignore the emergency supply capacity and buffering effect of hot / cold pipelines in disaster events, and on the other hand, lead to large-scale optimization problems and high solution complexity. It provides rules with simple form, strong interpretability, strong operability and safety, which can significantly improve the solution efficiency and have significant innovation and good application value.

[0273] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0274] A memory 1001 , a processor 1002 , and a computer program stored in the memory 1001 and executable on the processor 1002 .

[0275] When the processor 1002 executes the program, the security rule extraction and optimization method for improving the resilience of the urban energy system provided in the above embodiment is implemented.

[0276] Furthermore, the electronic device further includes:

[0277] The communication interface 1003 is used for communication between the memory 1001 and the processor 1002 .

[0278] The memory 1001 is used to store computer programs that can be run on the processor 1002 .

[0279] The memory 1001 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0280] If the memory 1001, the processor 1002, and the communication interface 1003 are implemented independently, the communication interface 1003, the memory 1001, and the processor 1002 can be connected to each other via a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 10 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0281] Optionally, in a specific implementation, if the memory 1001, the processor 1002 and the communication interface 1003 are integrated on a chip, the memory 1001, the processor 1002 and the communication interface 1003 can communicate with each other through an internal interface.

[0282] The processor 1002 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0283] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned security rule extraction and optimization method for improving the resilience of urban energy systems.

[0284] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0285] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0286] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0287] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array, a field programmable gate array, etc.

[0288] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0289] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A security rule extraction and optimization method for improving the resilience of urban energy systems, characterized in that: The following steps are involved: Calculating an urban energy system resilience index, and constructing a two-layer optimization model of the urban energy system based on the resilience index; Based on a preset energy transmission link, a preset energy storage link, and a preset energy conversion link, a dynamic characteristic model of a power electronics link of the urban energy system is modeled to obtain a dynamic characteristic model of a power electronics link; Modeling the multi-energy network dynamic characteristics of the urban energy system based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network, and the storage capacity of the district heating / cooling network pipelines to obtain a multi-energy network dynamic characteristics model; and Based on the dynamic characteristic model of the power electronics link and the dynamic characteristic model of the multi-energy network, a two-stage resilience operation model of the urban energy system under disaster events is constructed, and based on the preset sparse weight oblique decision tree model, the two-stage resilience model of the urban energy system under the disaster event is converted into a mixed integer linear constraint, and the mixed integer linear constraint is embedded in the two-layer optimization model of the urban energy system to obtain the optimization operation simulation results under the mixed integer linear constraint.

2. The method according to claim 1, characterized in that The calculation of urban energy system resilience indicators includes: The urban energy system resilience index is calculated based on a preset resilience index formula, wherein the preset resilience index formula is: ； in, is the toughness index, is the system performance before the disaster event, is the lowest value of system performance after a disaster event occurs, is the system load.

3. The method according to claim 2, characterized in that Before modeling the dynamic characteristics of the power electronics link of the urban energy system based on the preset energy transmission link, the preset energy storage link, and the preset energy conversion link to obtain the dynamic characteristics model of the power electronics link, the method further includes: The preset energy transmission link is determined based on a first instantaneous adjustment strategy of the response speed of the flexible soft switch, wherein the first instantaneous adjustment strategy is: ； ； in, for Time soft switching operation power, is a positive number, is the operating status of the distribution line, The upper limit of the operating power of the flexible soft switch; The preset energy storage link is determined based on a second instantaneous regulation strategy of the discharge power of the battery, wherein the second instantaneous regulation strategy is: ； in, is the operating power of the battery, is the upper limit of the battery's operating power, The time when the disaster occurred; The preset energy conversion link is determined based on a third instantaneous adjustment strategy of the input electrical power of the heat pump and the input electrical power of the compression refrigerator, wherein the third instantaneous adjustment strategy is: ； ； in, is the heat pump operating power, is the upper limit of the heat pump operating power, is the operating power of the compression refrigerator, is the upper limit of the operating power of the compression refrigerator, The time when the disaster occurred.

4. The method according to claim 3, characterized in that The multi-energy network dynamic characteristics model of the urban energy system is obtained by modeling the multi-energy network dynamic characteristics model based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network, and the storage capacity of the district heating / cooling network pipelines, including: Determining, based on the storage capacity of the natural gas grid and the storage capacity of the district heating / cooling network pipelines, a first relationship between the energy reserves of the pipeline network equivalent energy storage at the time of the disaster event and the operating state of the heat source at the time of the disaster event; Determining, based on the real-time balance of supply and demand of the power system, a second relationship between the energy reserve of the load thermal inertia equivalent energy storage at the time when the disaster event occurs and the heat / cooling load at the time when the disaster event occurs; The multi-energy network dynamic characteristics of the urban energy system are modeled according to the first relationship and the second relationship to obtain a multi-energy network dynamic characteristics model.

5. The method according to claim 4, characterized in that The objective function of the two-stage resilience operation model of the urban energy system is: ； in, is the load, For load The size of the load loss, is the load set, For time, is the time difference between two consecutive moments when the disaster occurs, The moment disaster restoration is completed.

6. The method according to claim 5, characterized in that The preset sparse weight oblique decision tree model is used to convert the two-stage resilience model of the urban energy system under the disaster event into mixed integer linear constraints, including: Obtaining a sparse weighted oblique decision tree based on the preset sparse weighted oblique decision tree model; The two-stage resilience model of the urban energy system under the disaster event is converted into a mixed integer linear constraint by utilizing the optimization objective function at each node of the sparse weight oblique decision tree.

7. The method according to claim 6, characterized in that The optimization objective function of the sparse weight oblique decision tree at each node is: ； in, is the optimal partition parameter vector, is the sample size, is the weight sum of the left child node, is the weight sum of the right child node, is the weighted information entropy of the left child node, is the weighted information entropy of the right child node, is the Elastic Net regularization term.

8. A security rule extraction and optimization device for improving the resilience of urban energy systems, characterized in that: include: A calculation module, configured to calculate an urban energy system resilience index and construct a two-layer optimization model of the urban energy system based on the resilience index; A generation module is used to model the dynamic characteristics of the power electronics link of the urban energy system based on a preset energy transmission link, a preset energy storage link, and a preset energy conversion link to obtain a dynamic characteristics model of the power electronics link; a modeling module for modeling the multi-energy network dynamic characteristics of the urban energy system based on the real-time balance of supply and demand of the power system, the storage capacity of the natural gas network, and the storage capacity of the district heating / cooling network pipelines, thereby obtaining a multi-energy network dynamic characteristics model; as well as An optimization module is used to construct a two-stage resilience operation model of the urban energy system under disaster events based on the dynamic characteristics model of the power electronic link and the dynamic characteristics model of the multi-energy network, and based on a preset sparse weight oblique decision tree model, convert the two-stage resilience model of the urban energy system under the disaster event into a mixed integer linear constraint, and embed the mixed integer linear constraint into a two-layer optimization model of the urban energy system to obtain the optimization operation simulation results under the mixed integer linear constraint.

9. The device according to claim 8, characterized in that The computing module is further configured to: The urban energy system resilience index is calculated based on a preset resilience index formula, wherein the preset resilience index formula is: ； in, is the toughness index, is the system performance before the disaster event, is the lowest value of system performance after a disaster event occurs, is the system load.

10. The device according to claim 9, characterized in that The generating module is further configured to: The preset energy transmission link is determined based on a first instantaneous adjustment strategy of the response speed of the flexible soft switch, wherein the first instantaneous adjustment strategy is: ； ； in, for Time soft switching operation power, is a positive number, is the operating status of the distribution line, The upper limit of the operating power of the flexible soft switch; The preset energy storage link is determined based on a second instantaneous regulation strategy of the discharge power of the battery, wherein the second instantaneous regulation strategy is: ； in, is the operating power of the battery, is the upper limit of the battery's operating power, The time when the disaster occurred; The preset energy conversion link is determined based on a third instantaneous adjustment strategy of the input electrical power of the heat pump and the input electrical power of the compression refrigerator, wherein the third instantaneous adjustment strategy is: ； ； in, is the heat pump operating power, is the upper limit of the heat pump operating power, is the operating power of the compression refrigerator, is the upper limit of the operating power of the compression refrigerator, The time when the disaster occurred.

11. The device according to claim 10, characterized in that The modeling module is further used to: Determining, based on the storage capacity of the natural gas grid and the storage capacity of the district heating / cooling network pipelines, a first relationship between the energy reserves of the pipeline network equivalent energy storage at the time of the disaster event and the operating state of the heat source at the time of the disaster event; Determining, based on the real-time balance of supply and demand of the power system, a second relationship between the energy reserve of the load thermal inertia equivalent energy storage at the time when the disaster event occurs and the heat / cooling load at the time when the disaster event occurs; The multi-energy network dynamic characteristics of the urban energy system are modeled according to the first relationship and the second relationship to obtain a multi-energy network dynamic characteristics model.

12. The device according to claim 11, characterized in that The objective function of the two-stage resilience operation model of the urban energy system is: ； in, is the load, For load The size of the load loss, is the load set, For time, is the time difference between two consecutive moments when the disaster occurs, The moment disaster restoration is completed.

13. The device according to claim 12, characterized in that The optimization module is further used to: Obtaining a sparse weighted oblique decision tree based on the preset sparse weighted oblique decision tree model; The two-stage resilience model of the urban energy system under the disaster event is converted into a mixed integer linear constraint by utilizing the optimization objective function at each node of the sparse weight oblique decision tree.

14. The device according to claim 13, characterized in that The optimization objective function of the sparse weight oblique decision tree at each node is: ； in, is the optimal partition parameter vector, is the sample size, is the weight sum of the left child node, is the weight sum of the right child node, is the weighted information entropy of the left child node, is the weighted information entropy of the right child node, is the Elastic Net regularization term.

15. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the security rule extraction and optimization method for improving the resilience of urban energy systems as described in any one of claims 1 to 7.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the security rule extraction and optimization method for improving the resilience of urban energy systems as described in any one of claims 1 to 7.

Images