Multi-energy power distribution system optimization planning method and device oriented to toughness improvement

Abstract

The invention provides a toughness improvement-oriented multi-energy power distribution system optimization planning method and device. The method comprises the following steps: acquiring a topological structure, equipment parameters, load data and historical operation data of the multi-energy power distribution system, and establishing a multi-energy power distribution system mathematical model comprising a power network, a thermal network and a gas network; the method comprises the following steps: identifying uncertain factors by using a multi-energy power distribution system mathematical model, and constructing a domain randomization sampling distribution function to form a random sample; establishing an uncertainty model of the multi-energy power distribution system according to the random sample, and obtaining system uncertainty characteristics through probability distribution characteristic analysis and statistical characteristic calculation; according to the uncertainty characteristics of the system, constructing a system toughness evaluation index system comprising a reliability index and a recovery capability index, and executing Monte Carlo simulation to form a system toughness evaluation result; and constructing a multi-objective optimization model by adopting a system toughness evaluation result. According to the invention, the toughness level of the multi-energy power distribution system is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system optimization planning, in particular to a multi-energy power distribution system optimization planning method and device for resilience improvement. BACKGROUND

[0002] As an important infrastructure of modern society, the reliability and resilience of the power system directly affect the stable operation of the social economy. With the large-scale access of distributed energy, energy storage systems and other new power equipment, the traditional distribution network is transforming into a modern multi-energy complementary power distribution system.

[0003] The common power distribution system planning methods at present mainly include two categories of deterministic planning and probabilistic planning. The deterministic planning method usually designs the system based on fixed load and power generation prediction, while the probabilistic planning considers the influence of uncertain factors and uses methods such as Monte Carlo simulation for optimization.

[0004] A typical existing technical solution adopts a hierarchical optimization structure, first performs power distribution network architecture planning in the upper layer to determine the type and location of main equipment, and then performs operation control strategy optimization in the lower layer. This solution introduces distributed energy and energy storage systems, and establishes a comprehensive energy system model including electricity, heat and natural gas and other forms of energy.

[0005] However, this solution has limitations in handling system uncertainty: first, the deterministic equivalent conversion method used has high computational complexity, making it difficult to handle large-scale systems; second, in the face of extreme weather and other emergencies, there is a lack of effective system reconfiguration and island operation strategy, and the system resilience is insufficient. SUMMARY

[0006] In view of this, the present application provides a multi-energy power distribution system optimization planning method and device for resilience improvement, which solves the problem of high complexity in handling system uncertainty and insufficient resilience in the prior art.

[0007] The embodiment of the present application provides a multi-energy power distribution system optimization planning method for resilience improvement, comprising the following steps:

[0008] Obtain the topology structure, device parameters, load data and historical operation data of the multi-energy power distribution system, and establish a multi-energy power distribution system mathematical model including the power network, the heat network and the gas network;

[0009] Use the multi-energy power distribution system mathematical model to identify uncertain factors and construct a domain randomization sampling distribution function to form a random sample;

[0010] According to the random sample, an uncertainty model of the multi-energy power distribution system is established, and the system uncertainty characteristics are obtained through probability distribution characteristic analysis and statistical characteristic calculation.

[0011] constructing a system resilience evaluation index system including reliability indexes and recovery capability indexes according to the system uncertainty characteristics, and performing Monte Carlo simulation to form a system resilience evaluation result;

[0012] constructing a multi-objective optimization model including investment cost, operation cost and system resilience by using the system resilience evaluation result;

[0013] solving the multi-objective optimization model by using a gradient-based domain randomization algorithm to obtain an optimized planning scheme of the multi-energy power distribution system;

[0014] forming a system reconfiguration strategy and island division scheme by identifying key nodes and weak links and designing strategies according to the optimized planning scheme of the multi-energy power distribution system.

[0015] Optionally, the domain randomization sampling distribution function is constructed to generate the random samples, including:

[0016] constructing a hybrid probability distribution model by using uncertainty factors in the multi-energy power distribution system, the hybrid probability distribution model including Beta distribution modeling for renewable energy output, normal distribution modeling for load fluctuation, and Weibull distribution modeling for device failure;

[0017] designing a random sampling strategy based on linear quadratic control by using the hybrid probability distribution model to generate a random sample set;

[0018] performing importance sampling and priority sampling on the random sample set to form the random samples.

[0019] Optionally, the system resilience evaluation index system is constructed according to the system uncertainty characteristics, the Monte Carlo simulation is performed, and the system resilience evaluation result is formed, including:

[0020] constructing reliability indexes including power supply reliability, system average interruption frequency index and system average interruption duration index, recovery capability indexes including fault isolation time, load recovery rate and system reconfiguration speed, and multi-energy complementary performance indexes including energy complementary index and energy conversion efficiency by using the system uncertainty characteristics to form an initial evaluation index set;

[0021] determining the weight of each index by using the analytic hierarchy process to generate a system resilience comprehensive evaluation index by using the initial evaluation index set;

[0022] performing the Monte Carlo simulation to comprehensively evaluate power supply reliability, system recovery capability and multi-energy complementary performance according to the system resilience comprehensive evaluation index to obtain the system resilience evaluation result.

[0023] Optionally, the system resilience evaluation result is used to construct a multi-objective optimization model including investment cost, operation cost and system resilience, including:

[0024] According to the system resilience evaluation result, an investment cost objective function including equipment investment cost, network transformation cost and control system cost, an operation cost objective function including energy purchase cost, equipment maintenance cost and system operation loss cost, and a resilience objective function including reliability index, recovery capability index and multi-energy complementarity index are constructed to form an objective function set;

[0025] The objective function set is used to set capacity constraints, network constraints, operation constraints and safety constraints to construct the multi-objective optimization model.

[0026] Optionally, the multi-objective optimization model is solved using a gradient-based domain randomization algorithm to obtain an optimized planning scheme of the multi-energy power distribution system, including:

[0027] Using the multi-objective optimization model, an initial parameter set of learning rate, iteration termination condition and sample number is initialized to form an algorithm initial parameter set;

[0028] According to the algorithm initial parameter set, a hybrid gradient calculation method with random disturbance is used to introduce momentum term and adaptive step strategy to generate gradient direction;

[0029] The gradient direction and the random sample are used to calculate the objective function value and the constraint violation degree to obtain an optimization evaluation result;

[0030] According to the optimization evaluation result, a double-layer nested iteration optimization of outer decision variable update and inner sample evaluation is performed to form the optimized planning scheme of the multi-energy power distribution system.

[0031] Optionally, according to the optimized planning scheme of the multi-energy power distribution system, a system reconstruction strategy is formed by identifying key nodes and weak links and designing strategies, including:

[0032] Using the key nodes and the weak links, a hierarchical structure including local reconstruction, regional reconstruction and global reconstruction is designed to form a hierarchical reconstruction strategy;

[0033] Using the hierarchical reconstruction strategy, specific measures including switch state adjustment, energy flow direction change, standby power supply start and energy conversion equipment operation mode adjustment are designed to construct a reconstruction measure scheme;

[0034] According to the reconstruction measure scheme, a reconstruction scheme feasibility verification is performed to form the system reconstruction strategy.

[0035] Optionally, the island partitioning scheme is designed according to the optimal planning scheme of the multi-energy power distribution system by identifying key nodes and weak links and designing strategies, and includes:

[0036] The candidate island regions are divided according to the self-sufficiency principle, the balance principle and the stability principle by using the multi-energy power distribution system mathematical model, and candidate island regions are formed;

[0037] The feasibility test result is obtained by using the candidate island regions to perform feasibility test on energy supply capacity, load distribution and multi-energy collaborative effect;

[0038] The preliminary island scheme is generated by determining the island boundary through an iterative optimization process according to the feasibility test result;

[0039] The island partitioning scheme is formed by performing stability analysis and supply-demand balance verification according to the preliminary island scheme.

[0040] Optionally, the multi-energy power distribution system mathematical model is established by including the power network, the heat network and the gas network, and includes:

[0041] The connection relationship among the power distribution network, the heat network and the gas network is established by using the topological structure, and a network connection relationship model is formed;

[0042] The characteristic model of the distributed power generation equipment, the energy storage device and the energy conversion equipment is constructed by combining the equipment parameters and the network connection relationship model, and an equipment mathematical model is generated;

[0043] The standardized data is obtained by using the load data to perform outlier processing and data standardization;

[0044] The multi-energy power distribution system mathematical model is constructed by integrating the standardized data, the network connection relationship model and the equipment mathematical model through model integration.

[0045] Optionally, it further includes verifying the optimal planning scheme of the multi-energy power distribution system, including:

[0046] The test conditions including typical working conditions and extreme working conditions are designed by using the optimal planning scheme, and a verification scenario is constructed;

[0047] The verification result is formed by using the verification scenario to perform static verification, dynamic verification and comprehensive verification;

[0048] The performance evaluation index is generated by evaluating the reconstruction time, the load recovery rate and the system stability margin according to the verification result, and the performance evaluation index is used to represent the feasibility of the optimal planning scheme.

[0049] Correspondingly, the embodiment of the application further provides a multi-energy power distribution system optimization planning device for resilience improvement, comprising:

[0050] a parameter collection module, configured to acquire a topology structure, equipment parameters, load data and historical operation data of the multi-energy power distribution system, and establish a multi-energy power distribution system mathematical model comprising a power network, a heat network and a gas network;

[0051] an uncertainty modeling module, configured to identify uncertainty factors by using the multi-energy power distribution system mathematical model, construct a domain randomization sampling distribution function, form a random sample, and establish an uncertainty model of the multi-energy power distribution system according to the random sample, and acquire system uncertainty characteristics;

[0052] a resilience evaluation module, configured to construct a system resilience evaluation index system comprising reliability indexes and recovery capability indexes according to the system uncertainty characteristics, and perform Monte Carlo simulation to form a system resilience evaluation result;

[0053] a model construction module, configured to construct a multi-objective optimization model comprising investment cost, operation cost and system resilience by using the system resilience evaluation result;

[0054] an optimization solution module, configured to solve the multi-objective optimization model by using a gradient-based domain randomization algorithm to obtain an optimization planning scheme of the multi-energy power distribution system;

[0055] a system reconstruction module, configured to form a system reconstruction strategy and an island division scheme by identifying key nodes and weak links and designing strategies according to the optimization planning scheme of the multi-energy power distribution system.

[0056] The embodiment of the application further provides a computer device, comprising at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the multi-energy power distribution system optimization planning method for resilience improvement.

[0057] The embodiment of the application further provides a computer readable storage medium storing computer instructions for causing a computer to perform the multi-energy power distribution system optimization planning method for resilience improvement.

[0058] The embodiment of the application further provides a computer program product comprising computer instructions, which, when executed by a processor, implement the steps of the multi-energy power distribution system optimization planning method for resilience improvement.

[0059] The application has the following technical effects:

[0060] The efficient modeling of system uncertainty is realized by introducing the domain randomization method, which reduces the computational complexity and improves the modeling efficiency;

[0061] A complete system resilience evaluation system is established to realize the quantitative evaluation of the system recovery capability;

[0062] An improved gradient-based domain randomization algorithm is proposed, which significantly improves the solving efficiency while ensuring convergence;

[0063] The system reconfiguration and island division strategy based on the optimization result is designed to improve the fault response capability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly introduced below. The drawings herein are incorporated into the specification and form a part of the specification, which show the embodiments consistent with the present disclosure, and are used to illustrate the technical solutions of the present disclosure together with the specification. It should be understood that the following drawings only show some embodiments of the present disclosure, and therefore should not be regarded as a limitation on the scope, and other related drawings can also be obtained by those skilled in the art without creative labor.

[0065] Figure 1 The flowchart of the multi-energy power distribution system optimization planning method for resilience improvement provided by the embodiments of the present application is shown.

[0066] Figure 2 The flowchart of the system reconfiguration strategy formed in the embodiments of the present application is shown.

[0067] Figure 3 The flowchart of the island division scheme designed in the embodiments of the present application is shown.

[0068] Figure 4 The structural diagram of the multi-energy power distribution system optimization planning device for resilience improvement in the embodiments of the present application is shown. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only some of the embodiments of the present disclosure, not all the embodiments. The components of the embodiments of the present disclosure described and shown in the drawings herein can be arranged and designed in various different configurations.

[0070] Therefore, the following detailed description of the embodiments of the present disclosure provided in the accompanying drawings is not intended to limit the scope of the claimed present disclosure, but merely represents selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present disclosure.

[0071] It should be noted that similar reference numbers and letters represent similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0072] The term "and / or" herein merely describes an association relationship, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0073] As Figure 1 shown, the embodiment of the present application provides a multi-energy power distribution system optimization planning method for resilience improvement, comprising the following steps:

[0074] S1: Obtain the topology structure, device parameters, load data and historical operation data of the multi-energy power distribution system, and establish a multi-energy power distribution system mathematical model including the power network, the heat network and the gas network.

[0075] Specifically, step S1 can include:

[0076] S1.1: Using the topology structure, establish the connection relationship of the power distribution network, the heat network and the gas network, and form a network connection relationship model.

[0077] First, the physical topology structure of the multi-energy power distribution system needs to be analyzed.

[0078] For the power network, the connection relationship of devices such as power distribution transformers, power distribution lines, circuit breakers, etc. needs to be determined; for the heat network, the arrangement of devices such as heat sources, heat pipes, heat exchange stations, etc. needs to be determined; for the gas network, the connection of devices such as gas sources, gas pipes, pressure regulating stations, etc. needs to be determined. Based on this physical structure information, a network connection matrix is established using graph theory to describe the topological relationship between different energy subsystems. In addition, the positions of energy conversion devices (such as combined heat and power units, gas turbines, etc.) also need to be marked, and these devices are the key nodes for realizing multi-energy complementation.

[0079] S1.2: Based on the device parameters and the network connection relationship model, construct the characteristic model of distributed power generation devices, energy storage devices, and energy conversion devices, and generate device mathematical models.

[0080] For distributed power generation devices, establish their output characteristics, start-stop characteristics, and operation constraint models. For example, photovoltaic power generation devices need to consider the relationship between light intensity and power generation power, and wind power generation needs to establish a wind speed-power curve model. For energy storage devices, focus on describing their charge-discharge characteristics, capacity constraints, and cycle life characteristics. For energy conversion devices, it is necessary to establish input-output characteristic models, such as the electric-thermal mutual supply characteristics of combined heat and power units and the conversion efficiency characteristics of electric-gas conversion devices. These device models need to be calibrated based on actual operating parameters and historical data to ensure the accuracy of the models.

[0081] S1.3: Use the load data to perform outlier processing and data standardization to obtain standardized data.

[0082] First, perform outlier detection and processing on the collected load data.

[0083] Use statistical methods (such as the 3σ criterion or boxplot method) to identify abnormal data points, and correct or remove outliers based on expert experience. Then, standardize different types of data to convert them to a unified numerical interval, facilitating subsequent modeling and analysis. The standardization process needs to consider the physical meaning of the data and choose appropriate standardization methods, such as maximum and minimum value standardization for periodic load data, and Z-score standardization for volatile data.

[0084] S1.4: Integrate the standardized data, the network connection relationship model, and the device mathematical model to construct the multi-energy power distribution system mathematical model through model integration.

[0085] In the final model integration step (step S1.4), the network connection relationship model, device mathematical model, and standardized data established earlier need to be integrated organically.

[0086] First, establish system state variables, including node voltage, power flow, and pipe network pressure; then build system operation constraint equations, including power balance constraints, network power flow constraints, and device operation constraints; finally, establish system dynamic characteristic equations to describe the dynamic response characteristics of the system at different time scales. During the model integration process, special attention is paid to maintaining the coordination between sub-models to ensure that the model accurately reflects the coupling relationship and complementary characteristics between multi-energy sources. In addition, considering the need for subsequent optimization calculations, attention needs to be paid to the normative description of the mathematical model and the feasibility of the solution during the model construction process.

[0087] So far, the embodiment of the present application establishes a complete multi-energy power distribution system mathematical model, which lays a foundation for subsequent uncertainty modeling and optimization planning.

[0088] The multi-energy power distribution system mathematical model has the advantages that the physical characteristics and operating characteristics of the multi-energy system are fully considered, a unified description of the static characteristics and dynamic characteristics of the system is achieved, and a reliable mathematical foundation is provided for the optimization planning of the system.

[0089] S2: Using the multi-energy power distribution system mathematical model, identifying uncertainty factors, and constructing a domain randomization sampling distribution function to form a random sample.

[0090] In the multi-energy power distribution system, there are three typical types of uncertainty factors. The first type is the uncertainty of renewable energy output. Taking photovoltaic power generation as an example, its output is affected by weather conditions, cloud cover, temperature changes and other factors. Assuming that a 100kW photovoltaic power station, its daytime power generation power presents a typical "hump-shaped" curve under sunny conditions, but in actual operation, it will appear random fluctuations. Through analysis of historical operation data, it can be found that its output distribution conforms to the Beta distribution characteristics, shape parameters α = 3.2, β = 2.8, this distribution can better describe the fluctuation characteristics of photovoltaic power generation.

[0091] The second type is the uncertainty of load demand. Taking the electricity load of a commercial area as an example, its basic load mode may be the peak of electricity consumption from 9:00 to 18:00 on weekdays, but the actual load will fluctuate due to weather, holidays, special activities and other factors. Analysis shows that the load fluctuation in a given period of time approximately obeys the normal distribution, the mean value is the predicted load value, and the standard deviation is about 10% of the mean value. For example, the predicted load of a certain period is 500kW, then the actual load may fluctuate within the distribution range of N(500, 50 2 ).

[0092] The third type is the uncertainty of equipment failure. Taking a transformer as an example, its failure time interval usually conforms to the Weibull distribution. Assuming that the historical operation data of a certain type of transformer shows that its characteristic life parameter λ = 8760 hours (i.e. 1 year), and the shape parameter k = 2.5, which indicates the characteristics of the device aging over time. Through this distribution model, the failure probability of the transformer at different operating periods can be predicted.

[0093] When constructing the domain randomization sampling, the correlation between these uncertainty factors needs to be considered. For example, in the summer high temperature weather, the photovoltaic power generation efficiency decreases at the same time, and the air conditioning load will increase significantly, which shows a negative correlation.

[0094] By introducing a Copula function to describe this correlation, more realistic random samples can be generated. Assuming a Gaussian Copula is used, with a correlation coefficient ρ = -0.6, the generated samples can reflect the negative correlation between photovoltaic output and load demand.

[0095] In the actual sampling process, a stratified sampling strategy is adopted to improve sampling efficiency. For example, the operating status of photovoltaic power generation is divided into three typical scenarios: "sunny day," "partly cloudy day," and "overcast day," and further refined sampling is performed under each scenario. Similarly, load demand is categorized and sampled according to "weekdays," "weekends," and "holidays." This stratified strategy ensures that the generated samples cover various typical operating conditions.

[0096] In this way, a random sample set containing various uncertainties can be generated.

[0097] For example, a set of samples containing 24-hour time series data is generated, including: time-of-use power values ​​of photovoltaic output, taking into account the fluctuation characteristics of weather effects; time-of-use power values ​​of load demand, reflecting the energy consumption characteristics of different periods; and possible equipment failure times. These samples will be used for subsequent system optimization and resilience assessment.

[0098] It should be noted that the number of samples generated needs to be determined based on the required computational precision and available computing resources. Generally, to ensure the reliability of statistical properties, the number of samples should not be less than 1000. Furthermore, the generated samples need to be validated to ensure that their statistical properties match the characteristics of the actual system.

[0099] Specifically, S2 includes:

[0100] S2.1: Utilize the uncertainties in the multi-energy power distribution system to construct a hybrid probability distribution model, which includes Beta distribution modeling for renewable energy output, normal distribution modeling for load fluctuations, and Weibull distribution modeling for equipment failures.

[0101] First, it is necessary to select an appropriate probability distribution model for different types of uncertainty factors.

[0102] To address the uncertainty in renewable energy output, a Beta distribution is used for modeling. This is because the Beta distribution has bounded and flexible shape characteristics, which can better describe the output characteristics of photovoltaic and wind power generation.

[0103] Specifically, by analyzing historical data to obtain the shape parameters α and β of the Beta distribution, the distribution function can accurately reflect the statistical characteristics of renewable energy output.

[0104] For the uncertainty of load fluctuation, considering that its randomness usually manifests as fluctuation around the mean, a normal distribution is adopted to model it, and the mean and standard deviation parameters are determined by the maximum likelihood estimation method. For the uncertainty of equipment failure, the Weibull distribution is adopted to describe the life characteristics and failure rate of the equipment, because it can well depict the time distribution characteristics of the equipment from the start of use to failure. By analyzing the historical failure data of the equipment, the scale parameter and shape parameter of the Weibull distribution are determined.

[0105] S2.2: Adopting the mixed probability distribution model, a random sampling strategy based on linear quadratic control is designed to generate a random sample set.

[0106] The state space model of the system is established, and the dynamic characteristics of the system are expressed in the form of state equations and output equations. Then, based on the linear quadratic control theory, a state feedback control law is constructed to enable the system to operate stably under random disturbances. On this basis, a random sampling strategy is designed: first, determine the sampling region, i.e., the value range of each uncertainty factor; then generate initial samples according to the aforementioned mixed probability distribution model; finally, adjust the samples through the linear quadratic controller to ensure that the generated samples not only meet the probability distribution characteristics but also satisfy the control constraints of the system. In particular, a covariance matrix is introduced during the sampling process to describe the correlation between different uncertainty factors, ensuring that the generated samples can reflect the coupling relationship between the components of the system.

[0107] S2.3: Perform importance sampling and priority sampling on the random sample set to form the random samples.

[0108] When performing importance sampling and priority sampling (step S2.3), the goal is to improve sampling efficiency and ensure the sufficiency of sampling for critical scenarios.

[0109] First, based on the requirements of system resilience assessment, identify the key scenarios that have a greater impact on system performance, such as renewable energy output fluctuations under extreme weather conditions, equipment failures during peak load periods, etc. Then, design an importance sampling strategy: construct an importance function so that the sampling probability is higher in the critical scenario region; use the acceptance-rejection sampling method to filter the initial samples according to the importance function, and retain representative samples. For priority sampling, based on system operation experience and historical data, establish a scenario priority evaluation mechanism to ensure that computing resources are prioritized for analyzing high-risk scenarios. In specific implementation, adopt hierarchical sampling technology to divide the sample space into multiple subspaces, and use different sampling densities for subspaces with different priorities.

[0110] Through the above steps, an efficient domain randomization sampling method is realized.

[0111] The innovation of the method is that the statistical characteristics of different types of uncertainties are accurately described by the mixed probability distribution model; the effectiveness of the sampling is guaranteed by combining the linear quadratic control theory; and the calculation efficiency is improved by importance sampling and priority sampling. Practice shows that compared with the traditional Monte Carlo sampling, the method can significantly reduce the number of samples required while ensuring the quality of the sampling, providing a high-quality random sample set for subsequent system optimization. It should be noted that the parameter settings of the sampling strategy need to be adjusted according to the characteristics of the specific system to balance the sampling efficiency and calculation accuracy.

[0112] S3: According to the random samples, an uncertainty model of the multi-capacity power distribution system is established, and the system uncertainty characteristics are obtained through probability distribution feature analysis and statistical characteristic calculation.

[0113] Based on the random samples obtained in step S2, this step aims to construct the uncertainty model of the system and quantify its characteristics.

[0114] Firstly, probability distribution feature analysis needs to be performed on the random samples. Specifically, for the random samples of renewable energy output, the mathematical statistical characteristics are calculated, including mean, variance, skewness and kurtosis, etc. These characteristic quantities can reflect the concentration trend and fluctuation characteristics of renewable energy output. For load fluctuation samples, in addition to calculating basic statistical quantities, the time series correlation needs to be analyzed, and the time-dependent characteristics of load fluctuation are described through autocorrelation function and cross-correlation function. For equipment failure samples, the focus is on analyzing the time interval distribution and duration distribution of failure occurrence, and constructing the failure probability density function.

[0115] Secondly, in order to accurately depict the overall uncertainty characteristics of the system, a multi-dimensional probability distribution model needs to be established. In addition, considering the complex correlation between various uncertainty factors in the system, Copula function is used to describe the dependence structure between multiple random variables.

[0116] In specific implementation, firstly, each edge distribution is mapped to the [0, 1] interval through probability integral transformation, and then the parameters of the Copula function are determined by maximum likelihood estimation method. The advantage of this method is that it can separate the edge distribution and the correlation structure, making the model more flexible and adaptable. In addition, in order to evaluate the influence of uncertainty on system performance, statistical characteristic calculation is needed. First, the statistical moments of the key performance indicators of the system (such as power supply reliability, energy utilization efficiency, etc.) are calculated through the moment method, including the expected value, variance, etc. Then, the non-parametric kernel density estimation method is used to construct the probability density function of the performance indicators. This method does not need to assume the distribution type in advance, and can better reflect the true distribution characteristics of the data. Especially for the coupling relationship between multiple performance indicators, the multi-dimensional kernel density estimation method is used for modeling.

[0117] In the model validation phase, the cross-validation method is used to evaluate the accuracy of the uncertainty model.

[0118] Specifically, the random sample set is divided into training set and test set, the model is established using the training set, and then the prediction effect of the model is verified on the test set. By calculating the prediction error and confidence interval, the reliability of the model can be quantified.

[0119] It should be noted that the complexity of the model should be balanced according to the actual demand. A too complex model may cause overfitting problem. Finally, based on the established uncertainty model, the key uncertainty characteristics of the system can be obtained, mainly including: (1) the probability distribution characteristics of various uncertainty factors, including marginal distribution and joint distribution information;

[0120] (2) the correlation structure between uncertainty factors, which describes the dependence relationship between multiple random variables; (3) the influence characteristics of uncertainty on system performance, including the statistical characteristics and risk measures of performance indicators. These characteristics provide an important theoretical basis for subsequent resilience assessment and optimization planning.

[0121] In summary, through probability distribution characteristic analysis and statistical characteristic calculation, a comprehensive system uncertainty model is established. This model not only can accurately describe the random characteristics of various uncertainty factors, but also can depict the complex correlation between them, laying a solid foundation for system resilience assessment.

[0122] S4: According to the system uncertainty characteristics, a system resilience evaluation index system containing reliability index and recovery capability index is constructed, and Monte Carlo simulation is performed to form the system resilience evaluation results.

[0123] Step S4 can include:

[0124] S4.1: Using the system uncertainty characteristics, a reliability index containing power supply reliability rate, system average interruption frequency index and system average interruption duration index, a recovery capability index containing fault isolation time, load recovery rate and system reconstruction speed, and a multi-energy complementary performance index containing energy complementary index and energy conversion efficiency are constructed to form an initial evaluation index set.

[0125] When constructing the initial evaluation index set (step S4.1), a complete index system needs to be established from the three dimensions of reliability, recovery capability and multi-energy complementarity.

[0126] Firstly, for reliability indicators, the availability of supply (ASAI) measures the continuous power supply capability of the system by calculating the ratio of the normal power supply time to the statistical period; the system average interruption frequency index (SAIFI) reflects the average number of power outages per user per unit time, which is calculated by counting the frequency of power outages; the system average interruption duration index (SAIDI) measures the average duration of power outages, which is obtained by dividing the total outage time of all users by the total number of users. The calculation of these indicators requires statistical analysis based on historical operation data and the uncertainty model established as described above.

[0127] Secondly, in terms of recovery capability indicators, fault isolation time reflects the speed of the system in identifying and isolating fault areas, which needs to consider the time consumption of fault location, protection action and switch operation; load recovery rate describes the system's ability to restore power supply after disturbance, which is calculated by integrating the ratio of recovered load to pre-fault load in the time dimension; system reconstruction speed measures the efficiency of adjusting the network topology and reconstructing the power supply path, which involves the generation and execution time of the reconstruction scheme. In particular, the characteristics of different types of faults and the dynamic process of system response need to be considered when calculating these indicators.

[0128] In terms of multi-energy complementary performance indicators, the energy complementary index evaluates the degree of multi-energy complementarity by calculating the collaborative energy supply effect between different energy subsystems, which needs to consider the temporal and spatial complementary characteristics and conversion efficiency of various types of energy; energy conversion efficiency focuses on the comprehensive energy utilization efficiency of energy conversion equipment (such as combined heat and power units, electric refrigeration equipment, etc.), which requires the establishment of detailed energy balance models and consideration of the dynamic characteristics of the equipment.

[0129] S4.2: Using the initial set of evaluation indicators, determine the weight of each indicator by the analytic hierarchy process (AHP) to generate a comprehensive evaluation index of system resilience.

[0130] In determining the weight of each indicator (step S4.2), the analytic hierarchy process (AHP) is used to establish a scientific weight distribution mechanism.

[0131] First, a hierarchical structure model is constructed, dividing the evaluation indicators into target layer, criterion layer and indicator layer; then a judgment matrix is established by expert scoring method, and the relative importance of each layer of indicators is calculated by eigenvalue method; finally, consistency test is conducted to ensure the rationality of weight distribution. It should be noted that the determination of weights should fully consider the actual operation characteristics of the system and user demand, and if necessary, fuzzy analytic hierarchy process can be introduced to handle the uncertainty of expert judgment.

[0132] S4.3: According to the system resilience comprehensive evaluation index, performing the Monte Carlo simulation, comprehensively evaluating the power supply reliability, system recovery ability and multi-energy complementary performance, and obtaining the system resilience evaluation result.

[0133] In the process of performing Monte Carlo simulation evaluation (step S4.3), it is necessary to design the simulation scene first, which includes normal operation scene and various fault scenes.

[0134] For each scene, random samples are generated based on the uncertainty model established in the foregoing, and then time sequence simulation analysis is performed. In the simulation process, the dynamic response characteristics of the system need to be considered, including the start-stop characteristics of the equipment, the dynamic characteristics of the network and the response characteristics of the control system. In particular, in order to improve the simulation efficiency, a hierarchical simulation strategy is adopted: the overall performance of the system is evaluated at the macro level, and the detailed response process under the key scene is analyzed at the micro level.

[0135] The processing of the simulation result is also an important link. First, the original simulation data is statistically analyzed, and the statistical characteristics of each evaluation index are calculated; then, based on the weights determined in the foregoing, a comprehensive evaluation is performed to obtain the quantitative evaluation result of the system resilience; finally, through sensitivity analysis, the key factors affecting the system resilience are identified. In addition, a visualization display mechanism of the evaluation result needs to be established to intuitively reflect the resilience level and potential risks of the system.

[0136] In summary, by establishing a multi-dimensional evaluation index system, combining scientific weight determination method and Monte Carlo simulation of the system, the overall evaluation of the resilience of the multi-energy distribution system is realized. The characteristics of this evaluation method are: not only considering the static reliability of the system, but also paying attention to the dynamic recovery ability, and also including the evaluation of multi-energy complementary performance, which provides a reliable decision basis for the optimization planning of the system.

[0137] S5: Using the system resilience evaluation result, a multi-objective optimization model including investment cost, operation cost and system resilience is constructed.

[0138] Step S5 can include:

[0139] S5.1: According to the system resilience evaluation result, an investment cost objective function including equipment investment cost, network transformation cost and control system cost, an operation cost objective function including energy purchase cost, equipment maintenance cost and system operation loss cost, and a resilience objective function including reliability index, recovery ability index and multi-energy complementary index are constructed to form an objective function set.

[0140] S5.2: Using the objective function set, capacity constraints, network constraints, operation constraints and safety constraints are set to construct the multi-objective optimization model.

[0141] When constructing the objective function for investment costs, the first consideration should be equipment investment costs. This includes the purchase costs of distributed generation equipment (such as photovoltaic and wind power generation equipment), energy storage devices (such as battery energy storage and thermal energy storage equipment), and energy conversion equipment (such as combined heat and power units and electro-gas conversion devices). These costs can typically be expressed as a piecewise linear function of equipment capacity to reflect economies of scale. Secondly, network upgrade costs mainly include the expenses for line upgrades, transformer replacements, and switchgear installations. These costs are directly related to changes in the network topology. Finally, control system costs include investments in monitoring equipment, communication equipment, and control software. This part of the cost needs to consider the system's level of intelligence and automation.

[0142] The objective function for investment cost can be expressed as:

[0143] F1 = Σ(αi×Ci_device + βj×Cj_network + γk×Ck_control)

[0144] Among them, α, β, and γ are the weighting coefficients of each cost item, which need to be determined based on actual engineering experience.

[0145] In constructing the objective function for operating costs, energy purchase costs are a key component, requiring consideration of market price fluctuations for various energy sources such as electricity and natural gas, as well as time-of-use pricing policies. Equipment maintenance costs are related to equipment operating time, start-up and shutdown frequency, and load levels, and a maintenance cost function based on an equipment lifespan model can be established. System operating loss costs mainly include network losses and energy conversion losses, which need to be determined through detailed power flow calculations and energy balance analysis.

[0146] The objective function for operating cost can be expressed as:

[0147] F2=Σ(λt×Pt_energy+μi×Mi_maintenance+νj×Lj_loss)

[0148] Where λ, μ, and ν are the corresponding cost coefficients, t represents the time period, and i and j represent the device and network numbers, respectively.

[0149] Regarding the resilience objective function, the aforementioned resilience assessment indices need to be integrated. Reliability indices mainly consider power supply reliability, average system outage frequency, and duration; recovery capability indices focus on fault isolation time, load recovery rate, and system reconfiguration speed; multi-energy complementarity indices include energy complementarity index and energy conversion efficiency.

[0150] The toughness objective function can be expressed as:

[0151] F3 = w1 × R reliability + w2 × R resilience + w3 × R multi-energy complementarity

[0152] where w1, w2, w3 are the weights of each type of index, and R represents the normalized resilience index.

[0153] In terms of constraint setting, firstly, there are capacity constraints, including the upper and lower limits of the capacity of various devices, the capacity limit of energy storage devices, etc. Secondly, there are network constraints, mainly including power flow constraints, voltage constraints, and pipeline pressure constraints. Operation constraints consider the start-stop characteristics of devices, climbing limits, and minimum operation time, etc. Safety constraints mainly focus on the stability margin of the system, standby capacity, and safe operation boundary. These constraints can be expressed as:

[0154] g(x)≤0 (inequality constraint)

[0155] h(x)=0 (equality constraint)

[0156] where x is the decision variable vector, including device capacity, network structure, and operation parameters, etc.

[0157] Finally, the multi-objective optimization model constructed can be expressed as:

[0158] min F=[F1,F2,-F3]

[0159] s.t.g(x)≤0;

[0160] h(x)=0;

[0161] x∈X.

[0162] It should be noted that since the resilience objective F3 is the larger the better, its negative value is taken in the optimization model. In addition, considering that there may be conflicts between the objective functions, a suitable multi-objective optimization algorithm is needed to solve the Pareto optimal solution set. The influence of uncertainty factors also needs to be considered in the solution process of the model, which can be handled by introducing a scenario set or probability constraint.

[0163] In summary, a multi-objective optimization model considering economic and resilience is constructed. The characteristics of this model are: it not only considers the long-term investment cost of the system, but also focuses on the operation cost and resilience index, and at the same time, through reasonable constraint conditions, the feasibility of the solution is guaranteed, providing a mathematical basis for the optimization planning of the system.

[0164] S6: A gradient-based domain randomization algorithm is used to solve the multi-objective optimization model and obtain the optimization planning scheme of the multi-energy power distribution system.

[0165] Step S6 can include: S6.1: Using the multi-objective optimization model, initialize the learning rate parameter, iteration termination condition, and sample number to form an initial parameter set of the algorithm.

[0166] The initial value setting of the learning rate parameter needs to balance the convergence speed and stability, and usually a small initial value is selected to ensure the stability of the algorithm. The iteration termination conditions include the maximum number of iterations, the relative change rate of the objective function value and the gradient norm, etc. The setting of these conditions needs to consider the balance between computational resources and optimization accuracy. The determination of the sample size needs to balance between computational efficiency and sampling sufficiency. A small sample size can be used in the initial stage, and the sample size can be gradually increased to improve the optimization accuracy as the optimization process progresses.

[0167] S6.2: According to the initial parameter set of the algorithm, a hybrid gradient calculation method with random disturbance is used, and a momentum term and an adaptive step size strategy are introduced to generate the gradient direction.

[0168] This method first applies random disturbance to the decision variables, and then calculates the gradient corresponding to each objective function. In order to improve the convergence performance of the algorithm, a momentum term is introduced to utilize historical gradient information, which helps the algorithm to jump out of local optimal solution and accelerate convergence. At the same time, an adaptive step size strategy is used to dynamically adjust the learning rate. When the objective function value continuously improves, the step size is increased to speed up convergence, and when oscillation occurs, the step size is reduced to improve stability. This adaptive mechanism can effectively balance the exploration ability and convergence speed of the algorithm.

[0169] S6.3: Using the gradient direction and the random sample, the objective function value and the constraint violation degree are calculated, and the optimization evaluation result is obtained.

[0170] First, the decision variables are updated according to the calculated gradient direction, and then the objective function value is calculated using random samples. At the same time, the constraint violation degree needs to be checked, and the solution that violates the constraint needs to be processed through the penalty function method. During the evaluation process, special attention is paid to the trade-off between different objectives, and the dominance relationship is calculated to determine whether the current solution can improve the Pareto front. In addition, the optimal solution set during the optimization process needs to be recorded, which will be used to generate the final optimization planning scheme.

[0171] S6.4: According to the optimization evaluation result, the double-layer nested iteration optimization of outer decision variable update and inner sample evaluation is performed to form the optimization planning scheme of the multi-energy power distribution system.

[0172] In the double-layer nested iterative optimization phase (step S6.4), an optimization strategy combining outer decision variable updating and inner sample evaluation is adopted. The outer iteration is responsible for updating decision variables, including device capacity, network structure, and other key parameters; the inner iteration, for evaluating the performance of each outer decision scheme in an uncertain environment through multiple random sampling. This double-layer structure can effectively handle the uncertainty in the optimization problem and improve the robustness of the solution. During the iteration process, an early stopping strategy is usually adopted to improve computational efficiency, i.e., when the inner evaluation results show that the current scheme is not competitive, the evaluation process of that scheme can be terminated early.

[0173] Finally, by recording the non-dominated solution set during the optimization process and combining the preference information of the decision maker, the most suitable solution is selected from the Pareto optimal solution set as the final optimization planning scheme. This scheme needs to meet the requirements of economy and resilience and have good implementation feasibility. In addition, sensitivity analysis of the final scheme is also needed to evaluate its robustness under different scenarios, providing a reference for the actual implementation of the scheme.

[0174] It is worth noting that the entire optimization process is an iterative optimization process, which needs to balance between computational efficiency and optimization quality. By reasonably setting algorithm parameters, using effective gradient calculation methods and evaluation strategies, the final optimization planning scheme that meets the actual demand can be obtained.

[0175] S7: According to the optimization planning scheme of the multi-capacity power distribution system, by identifying key nodes and weak links and designing strategies, system reconstruction strategies and island division schemes are formed.

[0176] As shown in Figure 2 , forming system reconstruction strategies includes:

[0177] S7.1: Using the key nodes and weak links, a hierarchical structure including local reconstruction, regional reconstruction, and global reconstruction is designed to form a hierarchical reconstruction strategy.

[0178] First, a three-level reconstruction hierarchy is established based on the physical structure and operating characteristics of the system.

[0179] Local reconstruction mainly targets single fault points or small-scale disturbances, and restores power supply by adjusting the network structure and device operating state around the fault point. This reconstruction method has fast response speed and small impact range, and is suitable for handling short-circuit faults of distribution lines, single device failures, etc. Regional reconstruction targets multiple fault points or medium-scale disturbances, and requires coordination of the operation of multiple power supply units and energy conversion devices.

[0180] This restructuring approach involves the coordinated efforts of multiple subsystems within a region, such as adjusting energy distribution and network topology in adjacent regions to support the affected area during a large-scale power outage in a certain region. Global restructuring is the highest level of response to systemic failures or large-scale disturbances, requiring the mobilization of resources throughout the system for coordinated restructuring. This restructuring approach is usually initiated during extreme weather events or major equipment failures, involving the re-planning of the overall system architecture.

[0181] S7.2: Adopting the hierarchical restructuring strategy, design specific measures including switch state adjustment, energy flow direction change, backup power supply startup, and energy conversion device operation mode adjustment to construct a restructuring measure scheme.

[0182] First is the switch state adjustment strategy, which restructures the power supply path by changing the state of circuit breakers and disconnectors, which requires considering the action timing of switch devices and the transient characteristics during load transfer. Second is the energy flow direction change strategy, which adjusts the energy transmission direction and power size between energy subsystems to achieve optimal energy distribution.

[0183] For example, during a power grid failure, the energy supply proportion of the gas network can be increased, or the discharge of the energy storage system can be used to supplement the energy supply gap. Third is the backup power supply startup strategy, including the rapid startup of distributed power generation equipment, the input of emergency generators, and the emergency connection with external power grids. Finally, the energy conversion device operation mode adjustment strategy changes the operation mode of combined heat and power units, electric-gas conversion devices, and other devices to improve the energy supply flexibility of the system.

[0184] S7.3: According to the restructuring measure scheme, perform the feasibility verification of the restructuring scheme to form the system restructuring strategy.

[0185] When performing the feasibility verification of the restructuring scheme (step S7.3), the feasibility of the restructuring scheme needs to be evaluated from multiple dimensions.

[0186] First, perform a technical feasibility analysis to check whether the reconstructed system meets the technical requirements such as voltage limit, power flow limit, device capacity constraint, etc.

[0187] Second, perform a timing feasibility analysis to ensure that the execution timing of each restructuring measure is reasonable and can complete the restructuring process within the specified time. Then, perform an economic feasibility analysis to evaluate the execution cost of the restructuring scheme and its impact on system operation efficiency.

[0188] Finally, perform a reliability analysis to verify whether the stability and power supply reliability of the reconstructed system meet the requirements.

[0189] It should be noted that the formulation of the reconstruction strategy needs to consider the dynamic characteristics of the system. After the fault occurs, the state of the system changes over time, and the reconstruction strategy needs to adapt to this dynamic change. Therefore, the reconstruction process usually adopts a step-by-step implementation method, and after each step of reconstruction is completed, the state is evaluated, and the next reconstruction action is determined according to the evaluation results. In addition, the reconstruction strategy also needs to consider the recovery priority of the system to ensure that important loads can be recovered first.

[0190] In summary, by establishing a hierarchical reconstruction structure, designing specific reconstruction measures, and conducting strict feasibility verification, a complete system reconstruction strategy is formed. The characteristics of this strategy are: clear hierarchy, specific measures, and rigorous verification, which can effectively guide the recovery and reconstruction process of the system after being disturbed. This strategy provides strong support for improving the resilience level of the system, enabling the system to maintain stable operation and quickly recover when facing various disturbances.

[0191] As shown in Figure 3 , the design of the island division scheme includes:

[0192] B1: Using the multi-energy power distribution system mathematical model, according to the self-sustaining principle, balance principle and stability principle, the region is divided to form a candidate island region.

[0193] In the region division stage, the system is partitioned based on three core principles. The self-sustaining principle requires that the divided island region has independent operation capability, and each region should contain sufficient power generation equipment, energy storage devices and necessary regulation equipment.

[0194] In practice, it is necessary to evaluate the installed capacity of distributed energy in the region, the capacity of energy storage system and the distribution of key loads. The balance principle emphasizes the balance between supply and demand among various island regions, requiring consideration of the uniformity of load distribution when dividing to avoid the situation that some regions have serious excess energy supply while other regions have serious shortage. The stability principle requires that each island region after division has sufficient regulation capability and control margin to maintain the stability of voltage and frequency within the region.

[0195] B2: Using the candidate island region, the feasibility of energy supply capacity, load distribution and multi-energy collaborative effect is verified, and the feasibility verification result is obtained.

[0196] In the feasibility verification stage, the energy supply capacity of each candidate island region needs to be evaluated first.

[0197] This includes the power generation capability of conventional power plants, the generation characteristics of renewable energy sources, and the charge-discharge capability of energy storage systems. At the same time, the load distribution characteristics need to be analyzed, including the temporal and spatial distribution of loads, energy use characteristics, and importance, etc. Special attention needs to be paid to the synergistic effects of multiple energy sources, such as the synergistic energy supply capability of combined heat and power systems, the complementary characteristics between different types of energy storage, etc. By establishing a detailed energy balance model, the supply-demand balance capability of each candidate region under different operating conditions is verified.

[0198] B3: According to the feasibility test results, determine the island boundary through an iterative optimization process to generate a preliminary island scheme.

[0199] In the stage of determining the island boundary, an iterative optimization method is used to gradually adjust and optimize the island boundary. First, the initial boundary is determined according to the physical connection relationship of the system, and then the boundary position is gradually optimized by evaluating the supply-demand characteristics, network characteristics, and control characteristics on both sides of the boundary. In the optimization process, special attention needs to be paid to the selection of the disconnection points at the boundary. These disconnection points should be controllable and should not cause stability problems after disconnection. At the same time, the impact of boundary adjustment on adjacent regions needs to be considered to ensure that each region still meets the self-sufficiency requirements after adjustment.

[0200] B4: According to the preliminary island scheme, perform stability analysis and supply-demand balance verification to form the island division scheme.

[0201] In the scheme verification stage, a comprehensive stability analysis and supply-demand balance verification need to be performed on the preliminary island scheme. Stability analysis includes transient stability and dynamic stability analysis, which needs to verify whether each island region can maintain stable operation when a disturbance occurs. This includes detailed analysis of voltage stability, frequency stability, and power angle stability. Supply-demand balance verification needs to consider balance problems at different time scales, including instantaneous power balance, daily energy balance, and seasonal energy balance. In particular, it needs to verify whether each island region can maintain reliable energy supply under renewable energy output fluctuations, load mutations, etc.

[0202] In addition, during the entire island division process, some practical factors also need to be considered. For example, the coverage and reliability of the communication system, which directly affects the coordinated control capability of island operation; the configuration requirements of the protection system to ensure effective fault protection in island mode; and the convenience of operation and maintenance, so that equipment maintenance and fault handling in island operation mode can be easily performed.

[0203] In summary, through systematic regional division, feasibility test, boundary optimization and scheme verification, a complete island division scheme is formed. This scheme not only considers technical feasibility, but also takes into account operation reliability and maintenance convenience, providing reliable technical support for the partition operation of the system in extreme cases. This scientific island division scheme is an important guarantee for improving the system resilience, which can effectively reduce the impact range of system failure and improve the fault recovery capability of the system.

[0204] In the embodiment of the application, the optimized planning scheme of the multi-energy power distribution system is verified, comprising:

[0205] A1: Using the optimized planning scheme, design test conditions including typical working conditions and extreme working conditions, and construct verification scenarios.

[0206] In terms of typical working conditions, it mainly includes daily operation scenarios, seasonal load change scenarios and equipment maintenance scenarios, etc. For example, the energy supply scenario during the summer peak load period is designed, considering factors such as photovoltaic power generation output fluctuation and user energy demand change; the multi-energy complementary scenario during the winter heating period is designed to verify the collaborative operation capability of the power grid, heat grid and gas grid. In terms of extreme working conditions, scenes such as severe weather conditions, serious equipment failure and emergencies are designed.

[0207] For example, simulate the scene of large-area equipment damage caused by typhoon weather to verify the emergency response capability of the system; design the scene of cascading failure of key equipment to test the fault isolation and load transfer capability of the system. The design of all verification scenarios needs to consider the actual operating conditions to ensure the authenticity and representativeness of the test.

[0208] A2: Using the verification scenarios, perform static verification, dynamic verification and comprehensive verification to form verification results.

[0209] Static verification mainly focuses on the performance indicators of the system in steady-state operation, including the inspection of parameters such as voltage distribution, power flow distribution and equipment load rate, to verify whether the system meets various technical constraints and operation requirements.

[0210] Dynamic verification focuses on testing the response characteristics of the system to disturbances, by applying load mutations, fault removal, equipment switching and other disturbances, observing the dynamic response process of the system, and evaluating the effectiveness of the control strategy and the stability performance of the system.

[0211] Comprehensive verification is carried out in more complex scenarios, such as multiple faults occurring simultaneously and multiple control targets conflicting with each other, to verify the comprehensive response capability of the system and the overall feasibility of the optimized scheme.

[0212] A3: According to the verification result, the reconstruction time, load recovery rate and system stability margin are evaluated to generate performance evaluation indexes, which are used to characterize the feasibility of the optimization planning scheme.

[0213] The evaluation of reconstruction time includes various links such as fault detection time, isolation time, reconstruction scheme generation time and execution time, and the average reconstruction time and maximum reconstruction time of the system are determined through statistical analysis.

[0214] The evaluation of load recovery rate needs to calculate the load recovery situation under different time scales, including the rapid recovery rate of important loads, the hierarchical recovery rate of general loads and the load recovery curve of the whole system.

[0215] The evaluation of system stability margin needs to analyze voltage stability margin, frequency stability margin and power angle stability margin, etc., to ensure that the system has sufficient safety margin under various operating conditions.

[0216] The analysis results of these performance evaluation indexes directly reflect the feasibility of the optimization planning scheme. For example, if the reconstruction time generally exceeds the expected target, it means that the reconstruction strategy needs to be optimized or the automation equipment needs to be increased; if the load recovery rate is not ideal, it may need to adjust the load partitioning scheme or increase the standby power supply capacity; if the stability margin is insufficient, it needs to consider increasing the regulating equipment or improving the control strategy. Through the comprehensive analysis of these indexes, the technical feasibility and economic rationality of the optimization planning scheme can be evaluated comprehensively.

[0217] It should be particularly noted that the performance evaluation process should use a combination of quantitative and qualitative methods. In addition to the above quantitative indexes, qualitative factors such as implementation difficulty, maintenance cost and future scalability also need to be considered. At the same time, the evaluation results should form a detailed analysis report to provide a basis for the improvement and optimization of the scheme. This system verification and evaluation method can effectively guarantee the reliability and practicality of the optimization planning scheme, and provide strong support for the actual implementation of the scheme.

[0218] In addition, as shown in Figure 4 The present application also provides a multi-energy power distribution system optimization planning device for resilience improvement, which includes a parameter acquisition module, an uncertainty modeling module, a resilience evaluation module, a model construction module, an optimization solving module and a system reconstruction module, and the specific functions of each module correspond to the corresponding steps in the above method. Specifically, it includes:

[0219] The parameter acquisition module is used to acquire the topology structure, device parameters, load data and historical operation data of the multi-energy power distribution system, and to establish a multi-energy power distribution system mathematical model including the power network, the heat network and the gas network;

[0220] An uncertainty modeling module is configured to identify uncertainty factors by using the multi-energy power distribution system mathematical model, construct a domain randomization sampling distribution function, form a random sample, and establish an uncertainty model of the multi-energy power distribution system according to the random sample, to obtain system uncertainty characteristics;

[0221] A resilience evaluation module is configured to construct a system resilience evaluation index system including reliability indexes and recovery capability indexes according to the system uncertainty characteristics, and perform Monte Carlo simulation to form a system resilience evaluation result.

[0222] A model construction module is configured to construct a multi-objective optimization model including investment cost, operation cost and system resilience by using the system resilience evaluation result.

[0223] An optimization solution module is configured to solve the multi-objective optimization model by using a gradient-based domain randomization algorithm to obtain an optimization planning scheme of the multi-energy power distribution system.

[0224] A system reconstruction module is configured to form a system reconstruction strategy and island division scheme by identifying key nodes and weak links and designing a strategy according to the optimization planning scheme of the multi-energy power distribution system.

[0225] The embodiments of the present disclosure further provide a computer readable storage medium having a computer program stored thereon, and the computer program is run by a processor to perform the steps of the multi-energy power distribution system optimization planning method for resilience improvement described in the method embodiments. The storage medium can be a volatile or non-volatile computer readable storage medium.

[0226] In addition, the embodiments of the present disclosure further provide a computer program product having a computer program stored thereon, and the computer program is run by a processor to perform the steps of the multi-energy power distribution system optimization planning method for resilience improvement provided by any of the embodiments of the present disclosure. For details, refer to the method embodiments described above, which will not be repeated here.

[0227] The computer program product can be specifically implemented by hardware, software or a combination thereof. In an optional embodiment, the computer program product is specifically embodied as a computer storage medium, which can be a volatile or non-volatile computer readable storage medium. In another optional embodiment, the computer program product is specifically embodied as a software product, such as a software development kit (Software Development Kit, SDK) and the like.

[0228] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the device and the apparatus described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here. In several embodiments provided in the present disclosure, it should be understood that the disclosed device, apparatus and method can be implemented in other ways. The apparatus embodiment described above is only schematic, for example, the division of the units is only a logical function division, and another division can be made in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0229] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0230] In addition, each functional unit in each embodiment of the present disclosure can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0231] If the functions are realized in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present disclosure or the part of the present disclosure that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present disclosure. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0232] Finally, it should be noted that the above-described embodiments are merely specific embodiments of the present disclosure, used to illustrate the technical solutions of the present disclosure, and are not intended to limit the present disclosure. The protection scope of the present disclosure is not limited thereto. Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can make modifications or easy changes to the technical solutions described in the foregoing embodiments, or easily think of changes or equivalent replacements for some of the technical features; and these modifications, changes or replacements do not cause the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A method for optimization planning of a multi-energy power distribution system oriented to resilience improvement, characterized in that, The method comprises the following steps: obtaining the topological structure, equipment parameters, load data and historical operation data of a multi-energy power distribution system, and establishing a multi-energy power distribution system mathematical model comprising a power network, a heat network and a gas network; using the multi-energy power distribution system mathematical model, identifying uncertain factors, and constructing a domain randomization sampling distribution function to form a random sample; establishing an uncertainty model of the multi-energy power distribution system according to the random sample, and obtaining system uncertainty characteristics through probability distribution characteristic analysis and statistical characteristic calculation; constructing a system resilience evaluation index system comprising reliability indexes and recovery capability indexes according to the system uncertainty characteristics, and performing Monte Carlo simulation to form system resilience evaluation results; using the system resilience evaluation results to construct a multi-objective optimization model comprising investment cost, operation cost and system resilience; using a gradient-based domain randomization algorithm to solve the multi-objective optimization model to obtain an optimized planning scheme of the multi-energy power distribution system; designing a system reconstruction strategy and island division scheme by identifying key nodes and weak links and designing strategies according to the optimized planning scheme of the multi-energy power distribution system.

2. The method of claim 1, wherein, The domain randomization sampling distribution function is constructed to generate the random sample, comprising: using the uncertain factors in the multi-energy power distribution system to construct a mixed probability distribution model, wherein the mixed probability distribution model comprises Beta distribution modeling for renewable energy output, normal distribution modeling for load fluctuation, and Weibull distribution modeling for equipment failure; using the mixed probability distribution model to design a random sampling strategy based on linear quadratic control to generate a random sample set; performing importance sampling and priority sampling on the random sample set to form the random sample.

3. The method of claim 1, wherein, The system resilience evaluation index system is constructed according to the system uncertainty characteristics, Monte Carlo simulation is performed, and system resilience evaluation results are formed, comprising: using the system uncertainty characteristics to construct a reliability index comprising a power supply reliability rate, a system average interruption frequency index and a system average interruption duration index, a recovery capability index comprising a fault isolation time, a load recovery rate and a system reconstruction speed, and a multi-energy complementary performance index comprising an energy complementary index and an energy conversion efficiency to form an initial evaluation index set; using the initial evaluation index set to determine the weight of each index by an analytic hierarchy process to generate a system resilience comprehensive evaluation index; performing the Monte Carlo simulation according to the system resilience comprehensive evaluation index to comprehensively evaluate the power supply reliability, the system recovery capability and the multi-energy complementary performance to obtain the system resilience evaluation results.

4. The method of claim 1, wherein, The multi-objective optimization model comprising investment cost, operation cost and system resilience is constructed using the system resilience evaluation results, comprising: constructing an investment cost objective function comprising equipment investment cost, network improvement cost and control system cost, an operation cost objective function comprising energy purchase cost, equipment maintenance cost and system operation loss cost, and a resilience objective function comprising reliability indexes, recovery capability indexes and multi-energy complementary indexes to form an objective function set according to the system resilience evaluation results; The target function set is used to set capacity constraints, network constraints, operation constraints and safety constraints to build the multi-objective optimization model.

5. The method of claim 1, wherein, The multi-objective optimization model is solved by using the gradient-based domain randomization algorithm to obtain an optimized planning scheme of the multi-energy power distribution system, including: The multi-objective optimization model is used to initialize learning rate parameters, iteration termination conditions and sample quantities to form an initial parameter set of the algorithm; According to the initial parameter set of the algorithm, a hybrid gradient calculation method with random disturbance is used to generate a gradient direction by introducing a momentum term and an adaptive step strategy; The gradient direction and the random sample are used to calculate the target function value and the constraint violation degree to obtain an optimization evaluation result; According to the optimization evaluation result, a double-layer nested iteration optimization of outer decision variable updating and inner sample evaluation is performed to form the optimized planning scheme of the multi-energy power distribution system.

6. The method of claim 1, wherein, According to the optimized planning scheme of the multi-energy power distribution system, a system reconstruction strategy is formed by identifying key nodes and weak links and designing strategies, including: The key nodes and the weak links are used to design a hierarchical structure including local reconstruction, regional reconstruction and global reconstruction to form a hierarchical reconstruction strategy; The hierarchical reconstruction strategy is used to design specific measures including switch state adjustment, energy flow direction change, backup power supply start and energy conversion device operation mode adjustment to build a reconstruction measure scheme; According to the reconstruction measure scheme, a reconstruction scheme feasibility verification is performed to form the system reconstruction strategy.

7. The method of claim 1, wherein, According to the optimized planning scheme of the multi-energy power distribution system, an island division scheme is designed by identifying key nodes and weak links and designing strategies, including: The multi-energy power distribution system mathematical model is used to divide the region according to the self-sustaining principle, the balance principle and the stability principle to form a candidate island region; The candidate island region is used to perform a feasibility test on energy supply capacity, load distribution and multi-energy synergy effect to obtain a feasibility test result; According to the feasibility test result, an island boundary is determined through an iterative optimization process to generate a preliminary island scheme; According to the preliminary island scheme, a stability analysis and supply-demand balance verification are performed to form the island division scheme.

8. The method of claim 1, wherein, The multi-energy power distribution system mathematical model is established, including: The topology structure is used to establish the connection relationship of the power distribution network, the heat network and the gas network to form a network connection relationship model; The device parameters and the network connection relationship model are combined to build characteristic models of distributed power generation equipment, energy storage devices and energy conversion devices to generate device mathematical models; The load data is used for outlier processing and data standardization to obtain standardized data; The standardized data, the network connection relationship model and the device mathematical model are integrated to build the multi-energy power distribution system mathematical model.

9. The method of claim 1, wherein, It also includes: The optimized planning scheme of the multi-energy power distribution system is verified, including: The optimized planning scheme is used to design test conditions including typical working conditions and extreme working conditions to build a verification scenario; The verification scene is used for performing static verification, dynamic verification and comprehensive verification to form a verification result; According to the verification result, reconstruction time, load recovery rate and system stability margin are evaluated to generate a performance evaluation index, which is used to represent the feasibility of the optimization planning scheme.

10. A device for optimization planning of a multi-energy power distribution system oriented to resilience enhancement, characterized in that, Comprise: A parameter acquisition module is configured to acquire the topological structure, equipment parameters, load data and historical operation data of the multi-energy power distribution system, and establish a multi-energy power distribution system mathematical model comprising a power network, a heat network and a gas network; An uncertainty modeling module is configured to identify uncertainty factors by using the multi-energy power distribution system mathematical model, construct a domain randomization sampling distribution function to form random samples, and establish an uncertainty model of the multi-energy power distribution system according to the random samples to obtain system uncertainty characteristics; A resilience evaluation module is configured to construct a system resilience evaluation index system comprising reliability indexes and recovery capability indexes according to the system uncertainty characteristics, and perform Monte Carlo simulation to form a system resilience evaluation result; A model construction module is configured to construct a multi-objective optimization model comprising investment cost, operation cost and system resilience by using the system resilience evaluation result; An optimization solving module is configured to solve the multi-objective optimization model by using a gradient-based domain randomization algorithm to obtain an optimization planning scheme of the multi-energy power distribution system; A system reconstruction module is configured to identify key nodes and weak links and design strategies according to the optimization planning scheme of the multi-energy power distribution system to form a system reconstruction strategy and an island division scheme.

Images