A comprehensive service system for intelligent design of electric power engineering and its use method

Abstract

The present invention discloses an integrated service system for intelligent design of electric power engineering and a method for using the system, which relates to the field of microgrid redundancy design and is used to solve the problem of dynamic resource configuration under load fluctuations and energy uncertainty. The system integrates multi-source data to construct a dynamic joint probability field model that can capture the temporal and spatial characteristics of load demand, distributed energy output and environmental variables, and comprehensively reflect their complex nonlinear dependencies. Based on the model, an advanced nonlinear optimization algorithm is used to derive the redundant capacity threshold in each time period, ensuring that the redundant configuration in different operating time periods can meet the high power supply reliability requirements while optimizing the economy and resource utilization efficiency. Furthermore, a Bayesian network is used to dynamically allocate redundant resources, fully tap the synergistic effect of multiple energy sources, and realize scientific allocation and real-time response of resources, thereby significantly enhancing the stability and economy of the microgrid in dealing with load fluctuations and energy uncertainty.

Description

Technical Field

[0001] The present invention relates to the field of microgrid redundancy design, and more specifically, to an integrated service system for intelligent design of electric power engineering and a method for using the system. Background Art

[0002] With the rapid development of distributed energy technologies, microgrids have become an important means of meeting local energy needs. By combining multiple energy sources (such as solar, wind, and battery storage), microgrids achieve flexible and efficient local power supply. However, to cope with unexpected failures or load fluctuations, a certain amount of redundant capacity is often reserved during the design phase to ensure power supply reliability. Reasonable planning of redundant capacity is a key step in microgrid design, directly impacting resource utilization, operational economy, and power supply security. In particular, in multi-energy synergy scenarios, achieving dynamic redundant configuration within limited resource constraints has become a core design challenge.

[0003] Traditional microgrid redundancy design is typically based on static load forecasts and empirical rules, making it difficult to fully reflect the probabilistic distribution characteristics of load fluctuations during actual operation. Especially when faced with the intermittent output of distributed energy resources, the randomness of load demand, and the complex influence of external environmental factors, static design solutions often suffer from insufficient redundancy or over-design. Insufficient redundancy can lead to reduced power supply reliability in emergencies and even cause local grid instability; while over-design can lead to waste of resources and increase initial investment and operating costs. In addition, existing methods rarely consider the dynamic changes in load fluctuations and failure risks in the time dimension, lacking the ability to accurately evaluate and optimize high-risk scenarios during critical periods. This design flaw not only limits the operational efficiency of microgrids but also increases design uncertainty and potential risks.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the Invention

[0005] To overcome the aforementioned shortcomings of the prior art, embodiments of the present invention provide a comprehensive service system and method for intelligent design of power engineering projects. By systematically integrating multi-source data, the present invention first constructs a dynamic joint probability field model that accurately captures the temporal and spatial characteristics of load demand, distributed energy output, and environmental variables, comprehensively reflecting their complex nonlinear dependencies. Based on this model, an advanced nonlinear optimization algorithm is employed to derive redundant capacity thresholds within each time period, ensuring that redundant configurations within different operating periods meet both high power supply reliability requirements and optimize economic efficiency and resource utilization. Furthermore, a Bayesian network is used to dynamically allocate redundant resources, fully exploiting the synergistic effects of multiple energy sources to achieve scientific resource allocation and real-time response. This significantly enhances the stability and economic efficiency of microgrids in responding to load fluctuations and energy uncertainty, effectively resolving the difficulty in dynamically balancing load and energy supply in traditional microgrid redundancy designs. Through multi-level and multi-dimensional optimization and collaboration, the system achieves comprehensive improvements in reliability, economic efficiency, and intelligence, significantly enhancing the overall adaptability and operational efficiency of microgrids in complex operating environments, and thus addressing the issues raised in the aforementioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for using a comprehensive service for intelligent design of electric power engineering, comprising the steps of:

[0008] S1. Construct a dynamic joint probability field model of load demand, distributed energy output, and environmental variables, capture time series dynamics through Markov chains, combine Copula functions to generate multi-scenario joint probability distributions, and identify the correlation and dynamic characteristics between variables.

[0009] S2. Based on the joint probability field model, the sensitive periods of load fluctuation and energy output are deduced, and the redundant capacity thresholds in different time periods are calculated using a dynamic programming algorithm to form a time-sensitive redundant configuration strategy.

[0010] S3. Based on the redundant capacity threshold, a Bayesian network is constructed to infer the energy node status, identify the synergistic relationship between various energy sources, and generate a dynamic redundant resource allocation plan to ensure the rational allocation of resources in multi-energy scenarios.

[0011] In a preferred embodiment, step S1 includes the following contents:

[0012] S1.1 First, through multi-source data collection and preprocessing, obtain real-time and historical data of various key parameters during the operation of the microgrid.

[0013] S1.2 In the time series dynamic characteristics modeling phase, the load demand changes are divided into several states, and the load demand transition probability matrix is ​​established using the Markov chain;

[0014] The modeling of distributed energy output uses an autoregressive hidden Markov model to define the hidden state and the corresponding output distribution, and trains the model parameters through the maximum expectation algorithm to characterize the intermittent and random nature of energy output.

[0015] The spatiotemporal modeling of environmental variables adopts the high Vickrickin interpolation method, which describes the correlation of environmental variables between different geographical locations by defining the spatial covariance matrix and combines it with the time series analysis method to achieve dynamic adjustment of environmental variables.

[0016] In a preferred embodiment, step S1 further includes the following:

[0017] S1.3 In the process of constructing the joint probability distribution, a multi-layer Copula function architecture is used to capture the nonlinear dependencies between load demand, distributed energy output, and environmental variables. First, the marginal distribution function F is defined for each type of data. L (L t ), F E (E t ), F V (V t ), corresponding to load demand, energy output and environmental variables respectively; then, a two-layer Copula structure is constructed. The first layer adopts a hybrid form of Gaussian Copula and Clayton Copula to capture the tail dependence and symmetry between different variables respectively; the second layer refines the dynamic dependency in the time series through conditional Copula; by introducing the time lag parameter τ, the joint probability distribution can reflect the dependency between the current moment and the past τ moments.

[0018] S1.4 In order to generate the joint probability distribution of multiple scenarios, the Markov chain Monte Carlo method is used to sample the constructed dynamic joint Copula model to generate a multi-scenario dataset covering different time periods and operating states.

[0019] The multi-scenario dataset generated by S1.5 is subjected to statistical tests to verify the accuracy and generalization ability of the model, ensuring that the generated multi-scenario joint probability distribution can truly reflect the dynamic relationship of multiple variables in the operation of the microgrid.

[0020] In a preferred embodiment, step S2 includes the following:

[0021] S2.1.1 First, divide the day into periods.

[0022] S2.1.2 Within each divided time period, use the dynamic joint probability field model to analyze the probability distribution characteristics of load demand and energy output, identify sensitive periods of high load fluctuations, low energy output, and extreme changes in environmental variables, and determine the degree of their impact on redundant capacity configuration by calculating the joint probability density of variables within each time period.

[0023] S2.2.1 In order to derive the redundant capacity thresholds for each time period, a redundant capacity demand model is constructed. By comprehensively analyzing the joint probability distribution of multiple variables, the optimal configuration range of redundant capacity, i.e., the redundant capacity demand R, is determined. t Specifically, the redundant capacity requirement R t At time t it is expressed as: in: represents the load demand change of the oth power node at time t; represents the output change of the oth distributed energy at time t; represents the change of the oth environmental variable at time t; N is the total number of power nodes, distributed energy resources and environmental variables.

[0024] S2.2.2 Based on the redundant capacity demand model, derive the redundant capacity threshold Θ for each time period t ; Use a multi-objective optimization framework to consider power supply reliability, economy and resource utilization efficiency to construct an optimization objective function in: represents the power supply reliability function; represents the economic function; represents the resource utilization efficiency function; λ1, λ2, λ3 are weight parameters that reflect the relative importance of each goal.

[0025] S2.2.3 During the optimization process, multi-level constraints and nonlinear optimization strategies are used to ensure the rationality and practicality of the redundant capacity threshold. The specific calculation formula is as follows: Where: P(L t,o ≤Θ t ) represents the probability that the load demand of the oth power node at time t does not exceed the redundancy threshold; C(Θ t ) is the cost function of redundant capacity, which is usually a nonlinear function and reflects the economic cost of redundant capacity; Indicates resource utilization efficiency and encourages efficient allocation of redundant capacity.

[0026] In a preferred embodiment, step S2 further includes the following:

[0027] S2.3.1 Based on the derived redundant capacity threshold, construct a time-sensitive redundant configuration strategy matrix S, whose elements S t,k It represents the configuration amount of the kth type of redundant resources in time period t, which is specifically expressed as: Where: T is the total number of divided time periods, and K is the number of categories of redundant resources.

[0028] S2.3.2 uses a method that combines hierarchical clustering and nonlinear optimization to generate a time-sensitive redundancy configuration strategy. The specific steps are as follows:

[0029] Hierarchical clustering: Based on the redundant capacity threshold and load demand fluctuation characteristics within each time period, a hierarchical clustering algorithm is used to divide the time period into several clusters. The time periods within each cluster have similar redundant capacity demand characteristics.

[0030] Non-linear optimization: For each cluster, a non-linear optimization algorithm is used to optimize the configuration ratio of redundant resources to ensure that the redundant capacity requirements of the cluster are met at all times while maximizing resource utilization efficiency and economy.

[0031] Strategy mapping: Maps the optimization results to the redundancy configuration strategy matrix, providing specific values ​​for the redundant resource configuration within each time period, ensuring that the redundancy configuration strategy is targeted and sensitive in different time periods.

[0032] In a preferred embodiment, step S3 includes the following contents:

[0033] S3.1 First, identify and define various redundant resource nodes in the microgrid, including backup generators, energy storage equipment, renewable energy devices, and critical load nodes.

[0034] S3.2 Based on the redundant resource node modeling, a Bayesian network is constructed to describe the conditional dependency relationship between nodes.

[0035] S3.3 reflects the nonlinear dependency between redundant resource nodes through a conditional probability table.

[0036] After completing the construction of the Bayesian network and its conditional probability table, S3.4 uses Bayesian reasoning technology and combines it with the redundant capacity threshold to optimize the allocation of dynamic redundant resources. The specific steps are as follows:

[0037] Observation data input and posterior probability calculation: The load demand L at the current moment t , energy output E t and environment variable V t As observation data, the Bayesian network is input and the posterior probability P(N i =s i|O), where O represents the set of observation data.

[0038] Energy synergy effect factor calculation: Energy synergy factor C is used ij , reflecting the synergistic effect between different energy types. The specific calculation formula is as follows: Where: ΔE i , ΔE j is the output change of energy types i and j at the current moment; is the attenuation parameter of the synergistic effect, which controls the attenuation rate of the synergistic effect with the output change difference; P(R i ) is the actual output power of energy type i; P max (R i ) is the maximum output power of energy type i.

[0039] S3.5 builds a multi-energy collaborative redundant resource allocation model with the goal of maximizing the overall power supply reliability and resource utilization efficiency of the system. The specific optimization objective function is defined as follows: Where: K is the number of redundant resource categories; S i 、S j is the configuration ratio of redundant resource categories i and j; is the energy synergy factor; P(N i =ON|O), P(N j =ON|O) is the posterior probability that nodes i and j are in the running state.

[0040] In a preferred embodiment, step S3 further includes the following:

[0041] S3.6 Under the defined objective function and constraints, solve the redundant resource allocation ratio S i The optimal value of , the specific optimization process includes:

[0042] Initialization: Generate the initial population or starting point.

[0043] Fitness evaluation: Calculate the fitness value of each configuration scheme, that is, the objective function The value of .

[0044] Selection and crossover: Select excellent individuals based on fitness, perform crossover and mutation operations, and generate a new generation of configuration solutions.

[0045] Iterative optimization: Repeat the fitness evaluation and selection crossover process until a preset convergence condition or number of iterations is reached.

[0046] Optimal solution extraction: The configuration solution with the highest fitness is finally selected as the optimal redundant resource allocation solution.

[0047] S3.7 generates a specific redundant resource allocation plan based on the optimization results, and organizes the optimal configuration ratios into a redundant resource configuration matrix, where each element represents the configuration ratio of the redundant resource category at the current moment.

[0048] A comprehensive service system for intelligent design of electric power engineering includes: a probability modeling module, a scenario generation module, a sensitive identification module, a threshold calculation module, a network reasoning module and a resource allocation module.

[0049] Probabilistic modeling module: Constructs a dynamic joint probability field model of load demand, distributed energy output and environmental variables, identifies the correlation and dynamic characteristics between variables, and generates a dynamic joint probability field model as the basis for the scenario generation module.

[0050] Scenario generation module: Uses Markov chain and Copula function to generate multi-scenario joint probability distribution from the probability model; the output multi-scenario joint probability distribution is used for analysis by the sensitive recognition module.

[0051] Sensitive identification module: Based on joint probability field model analysis, it deduces sensitive periods of load fluctuation and energy output; the identification results are input into the threshold calculation module to determine the redundant capacity threshold.

[0052] Threshold calculation module: uses dynamic programming algorithm to calculate the redundant capacity thresholds in different time periods; the output of the threshold calculation is sent to the network reasoning module for network reasoning and inferring the energy node status.

[0053] Network reasoning module: Constructs a Bayesian network to infer the status of energy nodes and identify the synergistic relationship between various energy sources; the inference results of network reasoning are used by the resource allocation module to generate dynamic allocation plans.

[0054] Resource allocation module: Generates dynamic redundant resource allocation plans based on redundant capacity thresholds and inference results to ensure reasonable allocation of resources in multi-energy scenarios.

[0055] The technical effects and advantages of the electric power engineering intelligent design integrated service system and its use method of the present invention are as follows:

[0056] By systematically integrating multi-source data, the present invention first constructs a dynamic joint probability field model that can capture the temporal and spatial characteristics of load demand, distributed energy output, and environmental variables, and comprehensively reflects their complex nonlinear dependencies. Based on this model, a nonlinear optimization algorithm is used to derive the redundant capacity thresholds within each time period, ensuring that the redundant configuration within different operating time periods can not only meet the high power supply reliability requirements but also optimize economic efficiency and resource utilization efficiency. Furthermore, a Bayesian network is used to dynamically allocate redundant resources, fully tap the synergistic effects of multiple energy sources, and achieve scientific resource allocation and real-time response. This significantly enhances the stability and economy of the microgrid in dealing with load fluctuations and energy uncertainties. It effectively solves the problem of dynamic balancing of load and energy supply in traditional microgrid redundancy design. Through multi-level and multi-dimensional optimization and coordination, it achieves a comprehensive improvement in system reliability, economy, and intelligence, and significantly enhances the overall adaptability and operational efficiency of the microgrid in complex operating environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 The present invention is a flowchart of a method for using integrated services for intelligent design of electric power engineering.

[0058] Figure 2 The figure is a structural diagram of an integrated service system for intelligent design of electric power engineering according to the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] Example 1: Figure 1 The present invention provides a method for using a comprehensive service for intelligent design of electric power engineering, comprising:

[0061] S1. Construct a dynamic joint probability field model of load demand, distributed energy output, and environmental variables, capture time series dynamics through Markov chains, combine Copula functions to generate multi-scenario joint probability distributions, and identify the correlation and dynamic characteristics between variables.

[0062] S2. Based on the joint probability field model, the sensitive periods of load fluctuation and energy output are deduced, and the redundant capacity thresholds in different time periods are calculated using a dynamic programming algorithm to form a time-sensitive redundant configuration strategy.

[0063] S3. Based on the redundant capacity threshold, a Bayesian network is constructed to infer the energy node status, identify the synergistic relationship between various energy sources, and generate a dynamic redundant resource allocation plan to ensure the rational allocation of resources in multi-energy scenarios.

[0064] In modern microgrid design, the dynamic changes in load demand, distributed energy output, and environmental variables have a crucial impact on system reliability and economic efficiency. Traditional static modeling methods struggle to fully capture the complex temporal and spatial dependencies between these variables, limiting the efficiency and effectiveness of redundancy configuration strategies. To address this issue, a dynamic joint probability field modeling approach is employed. This approach aims to comprehensively reflect the multivariable interactions in microgrid operation through advanced data processing techniques and complex probabilistic models, providing a scientific and precise probabilistic basis for redundancy configuration strategies.

[0065] Step S1 includes the following contents:

[0066] S1.1 First, through multi-source data collection and preprocessing, comprehensive real-time and historical data on various key parameters during microgrid operation is obtained. Load demand data includes load variations at different power nodes on different time scales; distributed energy output data covers the real-time power generation of renewable energy sources such as solar and wind power; and environmental variable data involves key environmental factors that affect energy output, such as temperature, humidity, wind speed, and irradiance. To ensure data quality and consistency, data cleaning techniques are used, including statistically based outlier detection and processing, spatiotemporal interpolation methods to fill missing data, and multivariate normalization to eliminate the influence of different dimensions on the model.

[0067] S1.2 In the time series dynamic characteristic modeling link, the load demand change is divided into several states (such as high load, medium load, and low load), and the load demand transition probability matrix T is established using the Markov chain. L , where T L (p1, p2) represents the probability of transitioning from state p1 to state p2. The transfer matrix is ​​estimated from historical load data using the maximum likelihood estimation method to ensure that the time series dynamic characteristics of load demand can be accurately reflected. The modeling of distributed energy output uses the autoregressive hidden Markov model (AR-HMM), defining the hidden state S E,t And the corresponding output distribution, the maximum expectation algorithm (EM algorithm) is used to train the model parameters to characterize the intermittent and random nature of energy output. The spatiotemporal modeling of environmental variables uses the high Vickrigian interpolation method (Kriging), by defining the spatial covariance matrix Σ V Describe the correlation between environmental variables in different geographical locations, and combine time series analysis methods to achieve dynamic adjustment of environmental variables.

[0068] S1.3 In the process of constructing the joint probability distribution, a multi-layer Copula function architecture is used to effectively capture the nonlinear dependencies between load demand, distributed energy output, and environmental variables. First, the marginal distribution function F is defined for each type of data. L (L t ), F E (E t ), F V (V t ), corresponding to load demand, energy output, and environmental variables, respectively. Next, a two-layer Copula structure is constructed. The first layer uses a hybrid of Gaussian Copula and Clayton Copula to capture tail dependencies and symmetry between different variables, respectively. The second layer further refines the dynamic dependencies in the time series through conditional Copula. Specifically, the joint Copula function is defined as:

[0069]

[0070] Among them, u L =F L (L t ),u E =F E (E t ),u V =F V (V t ) are the marginal cumulative distribution values ​​of load demand, energy output and environmental variables, respectively. and Represent the Gaussian Copula and Clayton Copula of the first layer respectively, and C2 is the mixed Copula of the second layer, which is used to integrate the output of the first layer and capture the dependency structure of higher levels. By introducing the time lag parameter τ, the joint probability distribution can reflect the dependency relationship between the current moment and the past τ moments, and the formula is expressed as:

[0071]

[0072] in, Represent the historical information of load demand, energy output and environmental variables at time t-τ respectively.

[0073] S1.4 In order to generate the multi-scenario joint probability distribution, the Markov Chain Monte Carlo (MCMC) method is used to sample the constructed dynamic joint Copula model to generate a multi-scenario dataset covering different time periods and operating states. The specific generation process is as follows:

[0074]

[0075] Where S is the number of generated scenes, usually S≥10 4 To ensure the adequacy and representativeness of the sample. represents the load demand, energy output and environmental variables under the sth scenario.

[0076] The multi-scenario dataset generated by S1.5 was statistically tested, including the Kolmogorov-Smirnov (KS) test and the Anderson-Darling (AD) test, to verify the agreement between the simulated joint probability distribution and the actual observed data:

[0077]

[0078] F sim (x) is the cumulative distribution function (CDF) of the simulated data.

[0079] F obs (x) is the cumulative distribution function of the observed data.

[0080] D KS is the Kolmogorov-Smirnov statistic, which indicates the maximum gap between the simulated distribution and the observed distribution.

[0081] A D The Anderson-Darling statistic evaluates the goodness of fit of the distribution by measuring the difference in weighted cumulative distribution functions.

[0082] n is the number of observation samples.

[0083] x i are the sorted observation data points.

[0084] Through these statistical tests, the accuracy and generalization ability of the model are verified, ensuring that the generated multi-scenario joint probability distribution can truly reflect the dynamic relationship of multiple variables in the operation of the microgrid.

[0085] Through these steps, the dynamic joint probability field model can accurately reflect the multivariable interaction dynamics in microgrid operation, providing a reliable data basis and theoretical support for the subsequent deduction of redundant capacity and the formulation of resource allocation strategies.

[0086] The core technical feature of this step is the introduction of a dynamic joint probability model that combines multi-layer Copula functions with time lag parameters. Traditional Copula methods usually only consider simple dependencies between variables, while this model captures dependencies at different levels by constructing a two-layer Copula structure, significantly improving the model's ability to express complex nonlinear relationships. In addition, by combining Markov chains and autoregressive hidden Markov models (AR-HMM), accurate time series modeling of load demand and distributed energy output is achieved, further enhancing the model's adaptability to dynamic changes. The high-Vickrigian interpolation method for modeling environmental variables ensures the continuity and accuracy of spatial characteristics in a dynamic environment.

[0087] The dynamic joint probability field model described above comprehensively and accurately simulates the dynamic interactions among load demand, distributed energy output, and environmental variables in a microgrid. This model not only improves the efficiency of data integration and processing but also significantly enhances the ability to predict microgrid operating conditions through a complex nonlinear dependency capture mechanism. The generated multi-scenario joint probability distribution covers a wide range of possible operating scenarios, providing a solid data foundation for subsequent redundant capacity deduction and resource allocation strategies, ensuring that the microgrid maintains efficient and reliable power supply capabilities in the face of changing environments and emergencies.

[0088] Through the implementation of this step, the dynamic balance problem in microgrid design can be fully supported and optimized at the data level, providing a scientific and accurate probabilistic basis for subsequent redundant configuration strategies, greatly improving the intelligence and adaptability of the microgrid system.

[0089] Based on the dynamic joint probability field model constructed in step S1, the dynamic characteristics of load demand, distributed energy output, and environmental variables in the microgrid system are fully reflected. However, an accurate probability model alone is not enough to directly guide the rational configuration of redundant capacity. It is necessary to further deduce redundant capacity thresholds for different time periods and formulate corresponding redundant configuration strategies to cope with the various complex scenarios and emergencies that the microgrid may face in actual operation.

[0090] Step S2 includes the following contents:

[0091] S2.1.1 First, divide the day into multiple time periods. Based on the dynamic characteristics of load demand and energy output, a 24-hour period can generally be divided into different phases, such as the morning peak, the midday trough, the evening peak, and the nighttime stable period. Specific division criteria can be based on statistical analysis of historical load and energy output data to identify key periods with large fluctuations in load demand and energy output.

[0092] S2.1.2 Within each defined time period, use a dynamic joint probability field model to analyze the probability distribution characteristics of load demand and energy output, identifying periods of high load fluctuation, low energy output, and extreme changes in environmental variables. By calculating the joint probability density of the variables within each time period, determine their impact on the redundant capacity configuration.

[0093] S2.2.1 In order to accurately derive the redundant capacity thresholds for each time period, a redundant capacity demand model is constructed. This model takes into account the volatility of load demand, the uncertainty of energy output, and the impact of environmental variables. By comprehensively analyzing the joint probability distribution of multiple variables, the optimal configuration range of redundant capacity is determined, namely, the redundant capacity demand R t Specifically, the redundant capacity requirement R t At time t it can be expressed as:

[0094]

[0095] in:

[0096] It represents the load demand change of the oth power node at time t.

[0097] Represents the output change of the oth distributed energy at time t.

[0098] Represents the change of the oth environmental variable at time t.

[0099] N is the total number of power consumption nodes, distributed energy resources and environmental variables.

[0100] By integrating the changes in various variables, the demand for redundant capacity at time t is quantified.

[0101] S2.2.2 uses an adaptive dynamic optimization algorithm to derive the redundant capacity threshold Θ for each time period based on the redundant capacity demand model. t Using a multi-objective optimization framework, considering power supply reliability, economy and resource utilization efficiency, the optimization objective function is constructed.

[0102]

[0103] in:

[0104] Represents the power supply reliability function, ensuring that the redundant capacity can meet the power supply demand during high load fluctuations.

[0105] Represents the economic function that minimizes the initial investment and operating costs of the redundant configuration.

[0106] It represents the resource utilization efficiency function, optimizes the allocation of redundant capacity, and avoids resource waste.

[0107] λ1, λ2, and λ3 are weight parameters that reflect the relative importance of each target.

[0108] The power supply reliability function aims to quantify the contribution of redundant capacity to system power reliability. It calculates the probability of the system meeting power demand during high load fluctuations at different redundant capacity thresholds, thereby ensuring the continuity and stability of power supply.

[0109]

[0110] Where: P(L t,o ≤Θ t ) indicates that the load demand of the oth power node at time t does not exceed the redundancy threshold Θ t The probability of . N is the total number of power nodes.

[0111] The overall power supply reliability of the system is reflected by averaging the probability of each node meeting the power supply demand. The goal is to maximize this probability, that is, to minimize the power supply reliability function.

[0112] The economics function is used to quantify the economic cost of redundant configurations. This function not only considers the initial investment cost but also includes operating and maintenance costs to ensure that redundant configurations are economically feasible.

[0113]

[0114] Where: C init (S o ) represents the oth type of redundant resource S o The initial investment cost. op (S o ) represents the oth type of redundant resource S o The operation and maintenance costs. β is the nonlinear growth coefficient, reflecting the exponential growth characteristics of the initial cost as the redundant capacity threshold increases. ln(Θ t +1) is the logarithmic function of the redundant capacity threshold, indicating that as the redundant capacity increases, the increasing rate of the operating cost gradually slows down.

[0115] This function accurately reflects the impact of redundant capacity on economic costs through a nonlinear relationship, avoiding the limitations of simple linear superposition and ensuring the accuracy of economic evaluation.

[0116] The resource utilization efficiency function aims to optimize the allocation of redundant capacity, improve resource utilization efficiency, and avoid over-allocation and waste of resources.

[0117]

[0118] in:

[0119] Indicates the sum of all redundant capacity thresholds, which serves as the benchmark for resource utilization.

[0120] Indicates the ratio of the redundant capacity threshold to the demand, ensuring that the redundant capacity is not over-provisioned.

[0121] γ is a nonlinear adjustment parameter that controls the sensitivity and responsiveness of the resource utilization efficiency function.

[0122] This function nonlinearly measures resource utilization efficiency by raising the power of the ratio of redundant capacity to demand. It encourages minimizing the configuration of redundant capacity and improving resource utilization while meeting demand.

[0123] S2.2.3 During the optimization process, multi-level constraints and nonlinear optimization strategies are used to ensure the rationality and practicality of the redundant capacity threshold. The specific calculation formula is as follows:

[0124]

[0125] in:

[0126] P(L t,o ≤Θ t ) represents the probability that the load demand of the oth power node at time t does not exceed the redundancy threshold.

[0127] C(Θ t ) is the cost function of redundant capacity, which is usually a nonlinear function, such as an exponential function or a logarithmic function, reflecting the economic cost of redundant capacity.

[0128] Indicates resource utilization efficiency and encourages efficient allocation of redundant capacity.

[0129] By solving the above optimization problem, the redundant capacity thresholds for each time period are obtained, ensuring that the redundant capacity configuration in different time periods can not only meet the power supply reliability requirements, but also control economic costs and improve resource utilization efficiency.

[0130] S2.3.1 Based on the derived redundant capacity threshold, construct a time-sensitive redundant configuration strategy matrix S, whose elements S t,k It represents the configuration amount of the kth type of redundant resources (such as backup generators, energy storage equipment, etc.) in time period t. It is specifically expressed as:

[0131]

[0132] Where: T is the total number of time periods divided. K is the number of categories of redundant resources.

[0133] S2.3.2 uses a combination of hierarchical clustering and nonlinear optimization to generate a time-sensitive redundancy configuration strategy. The specific steps are as follows:

[0134] Hierarchical clustering: Based on the redundant capacity threshold and load demand fluctuation characteristics within each time period, a hierarchical clustering algorithm (such as Ward's method) is used to divide the time period into several clusters. The time periods within each cluster have similar redundant capacity demand characteristics.

[0135] Nonlinear optimization: For each cluster, a nonlinear optimization algorithm (such as particle swarm optimization, genetic algorithm) is used to optimize the configuration ratio S of redundant resources. t,k , ensuring that redundant capacity requirements are met at all times within the cluster while maximizing resource utilization efficiency and economy.

[0136] Strategy mapping: Maps the optimization results to the redundancy configuration strategy matrix, providing specific values ​​for the redundant resource configuration within each time period, ensuring that the redundancy configuration strategy is targeted and sensitive in different time periods.

[0137] By developing a time-sensitive redundancy configuration strategy, microgrid design can dynamically adjust redundant capacity within different time periods, ensuring high power supply reliability during periods of high load fluctuations and low energy output. It also optimizes resource utilization during periods of abundant resources and reduces the economic cost of redundant configuration. This not only improves the overall operational efficiency and cost-effectiveness of the microgrid, but also enhances its resilience to uncertainty and emergencies.

[0138] In step S2, a time-sensitive redundancy configuration strategy was successfully derived, clarifying the redundant capacity requirements within different time periods. However, how to scientifically and rationally allocate these redundant resources in a multi-energy synergy scenario remains a key issue for ensuring efficient microgrid operation. Traditional resource allocation methods often ignore the synergy between multiple energy sources, resulting in inefficient resource utilization and insufficient power supply reliability. Therefore, it is necessary to introduce a more advanced Bayesian network model, incorporating synergy factors, to address the complexity of multi-energy synergy resource allocation.

[0139] In microgrid redundancy design, the coordinated operation of multiple energy sources is a key factor in improving system reliability and resource utilization efficiency. Building on the time-sensitive redundancy configuration strategy derived in the previous step, this step aims to dynamically allocate redundant capacity thresholds by constructing and applying a Bayesian network, generating a redundant resource allocation scheme suitable for multi-energy coordination scenarios. This process involves modeling redundant resource nodes, constructing a Bayesian network structure, generating conditional probability tables, and generating a dynamic optimization allocation strategy to ensure that redundant resources are scientifically and rationally configured in complex and changing operating environments.

[0140] Step S3 includes the following contents:

[0141] S3.1 First, identify and define various redundant resource nodes in the microgrid, including backup generators, energy storage equipment, renewable energy devices (such as solar panels and wind turbines), and critical load nodes. Each redundant resource node N i It has unique attributes and state variables to describe its operating characteristics and availability. For example, the backup generator node G i The state variables of the energy storage device node B can be defined as "run" (ON) and "stop" (OFF); j The state variables of can be defined as “charging”, “discharging” and “idle”; the renewable energy node R k The state variables include "high output", "medium output" and "low output". By systematically modeling these nodes, a complex network structure is formed, reflecting the dependencies and synergies between the resource nodes.

[0142] S3.2 Based on the redundant resource node modeling, a Bayesian Network (BN) is constructed to describe the conditional dependencies between nodes. A Bayesian Network consists of nodes and directed edges, where each node represents a redundant resource or load node, and the edge represents the conditional dependencies between nodes. For example, the energy storage device node B j The state of the backup generator node G may depend on i The operating status of the renewable energy node R k This network structure can capture the complex dynamic interactions between resource nodes in a multi-energy system. The Bayesian network's structural design must be based on the microgrid's actual operating data and resource characteristics to ensure the model accurately reflects the system's true operating status.

[0143] S3.3 reflects the nonlinear dependencies between redundant resource nodes through a conditional probability table. Traditional CPT typically uses a simple conditional probability distribution, while this invention introduces a multi-level, nonlinear conditional probability formula to enhance the model's expressiveness. Specifically, the following conditional probability formula is used:

[0144]

[0145] in:

[0146] N i is the target node.

[0147] s i For node N i status.

[0148] Pa(N i ) is node N iThe parent node collection.

[0149] pa is the specific status of the parent node.

[0150] M is the number of characteristic functions.

[0151] φ j is the characteristic function f j The associated weight parameter.

[0152] f j (s i ,pa) is the characteristic function that defines the nonlinear relationship between the target node state and the parent node state.

[0153] Through the multi-level feature function f j , such as interactive characteristic functions and nonlinear cumulative characteristic functions, can capture the complex synergistic effects between different redundant resources. For example, the following characteristic functions are defined:

[0154]

[0155] in:

[0156] P(R k ) is renewable energy R k Output power.

[0157] P max (R k ) is renewable energy R k Maximum output power.

[0158] I is the indicator function, when the backup generator G i The value is 1 when in running state, otherwise it is 0.

[0159] Θ j Energy storage device B corresponding to the redundant capacity threshold j configuration ratio.

[0160] V t is the value of the environmental variable at time t (such as temperature).

[0161] V opt for the best working value of the environment variable.

[0162] σ V The standard deviation of the variation in the environmental variable.

[0163] These characteristic functions enhance the Bayesian network's ability to model multi-energy synergy effects by introducing nonlinear transformations and interactions, ensuring that the redundant resource allocation scheme can fully consider the dynamic interactions between various energy types.

[0164] After completing the construction of the Bayesian network and its conditional probability table, S3.4 uses Bayesian reasoning technology and combines it with the redundant capacity threshold to optimize the allocation of dynamic redundant resources. The specific steps are as follows:

[0165] Observation data input and posterior probability calculation: The load demand L at the current moment t , energy output E t and environment variable V t As observation data, the Bayesian network is input and the posterior probability P(N i =s i |O), where O represents the set of observation data.

[0166] Energy synergy effect factor calculation: Energy synergy factor C is used ij , reflecting the synergistic effect between different energy types. The specific calculation formula is as follows:

[0167]

[0168] in:

[0169] ΔE i , ΔE j is the output change of energy types i and j at the current moment.

[0170] is the attenuation parameter of the synergistic effect, which controls the attenuation rate of the synergistic effect with the output change difference.

[0171] P(R i ) is the actual output power of energy type i.

[0172] P max (R i ) is the maximum output power of energy type i.

[0173] This formula calculates the energy synergy factor by combining the output change difference and the current output level to dynamically reflect the synergistic relationship between different energy types.

[0174] S3.5 builds a multi-energy collaborative redundant resource allocation model with the goal of maximizing the overall power supply reliability and resource utilization efficiency of the system. The specific optimization objective function is defined as follows:

[0175]

[0176] in:

[0177] K is the number of redundant resource categories.

[0178] S i 、S jis the configuration ratio of redundant resource categories i and j.

[0179] It is an energy synergistic factor.

[0180] P(N i =ON|O), P(N j =ON|O) is the posterior probability that nodes i and j are in the running state.

[0181] This objective function optimizes the overall power supply reliability and resource utilization efficiency of the system by maximizing the product of the redundant resource configuration ratio and the energy synergy factor, comprehensively considering the synergy effects and operating state probabilities among various resource types.

[0182] S3.6 uses a nonlinear optimization algorithm (such as genetic algorithm, simulated annealing algorithm or particle swarm optimization algorithm) to solve the redundant resource allocation ratio S under the defined objective function and constraints. i The specific optimization process includes:

[0183] Initialization: Generate the initial population or starting point.

[0184] Fitness evaluation: Calculate the fitness value of each configuration scheme, that is, the objective function The value of .

[0185] Selection and crossover: Select excellent individuals based on fitness, perform crossover and mutation operations, and generate a new generation of configuration solutions.

[0186] Iterative optimization: Repeat the fitness evaluation and selection crossover process until a preset convergence condition or number of iterations is reached.

[0187] Optimal solution extraction: The configuration solution with the highest fitness is finally selected as the optimal redundant resource allocation solution.

[0188] Through nonlinear optimization algorithms, the redundant resource configuration scheme is ensured to achieve the optimal balance between power supply reliability and resource utilization efficiency in multi-energy synergy scenarios.

[0189] S3.7 generates a specific redundant resource allocation plan based on the optimization results. i A redundant resource allocation matrix is ​​constructed, where each element represents the current allocation ratio of a redundant resource category. This allocation plan is transmitted to the actual resource management system through the system control module, guiding the dynamic allocation and operation of redundant resources, ensuring the efficient and stable operation of the microgrid in different energy synergy scenarios.

[0190] By implementing this step, microgrids can achieve scientific and dynamic allocation of redundant resources in multi-energy synergy scenarios, significantly improving power supply reliability and resource utilization efficiency. Specifically, during periods of high load fluctuations and low energy output, redundant resources can respond quickly to ensure power supply continuity. During periods of abundant resources, optimized allocation reduces the economic cost of redundant resources and avoids resource waste. Furthermore, the introduction of multi-energy synergy enhances the microgrid's ability to cope with complex operating environments and emergencies, improving the overall resilience and intelligence of the system.

[0191] Example 2: Figure 2 The present invention provides an integrated service system for intelligent design of electric power engineering, which includes a probability modeling module, a scenario generation module, a sensitive identification module, a threshold calculation module, a network reasoning module and a resource allocation module.

[0192] Probabilistic modeling module: Constructs a dynamic joint probability field model of load demand, distributed energy output and environmental variables, identifies the correlation and dynamic characteristics between variables, and generates a dynamic joint probability field model as the basis for the scenario generation module.

[0193] Scenario generation module: Uses Markov chain and Copula function to generate multi-scenario joint probability distribution from the probability model; the output multi-scenario joint probability distribution is used for analysis by the sensitive recognition module.

[0194] Sensitive identification module: Based on joint probability field model analysis, it deduces sensitive periods of load fluctuation and energy output; the identification results are input into the threshold calculation module to determine the redundant capacity threshold.

[0195] Threshold calculation module: uses dynamic programming algorithm to calculate the redundant capacity thresholds in different time periods; the output of the threshold calculation is sent to the network reasoning module for network reasoning and inferring the energy node status.

[0196] Network reasoning module: Constructs a Bayesian network to infer the status of energy nodes and identify the synergistic relationship between various energy sources; the inference results of network reasoning are used by the resource allocation module to generate dynamic allocation plans.

[0197] Resource allocation module: Generates dynamic redundant resource allocation plans based on redundant capacity thresholds and inference results to ensure reasonable allocation of resources in multi-energy scenarios.

[0198] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0199] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0200] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article or device. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0201] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for using integrated services for intelligent design of electric power engineering, characterized in that: include: S1. Construct a dynamic joint probability field model of load demand, distributed energy output, and environmental variables. This model captures temporal dynamics through Markov chains and combines Copula functions to generate multi-scenario joint probability distributions, identifying the correlations and dynamic characteristics between variables. S2. Based on the joint probability field model, we deduce sensitive periods of load fluctuation and energy output, and use a dynamic programming algorithm to calculate the redundant capacity thresholds for different time periods to develop a time-sensitive redundant configuration strategy. Step S2 includes the following contents: S2.1.1 First, divide the day into multiple time periods; S2.1.2 Within each divided time period, use a dynamic joint probability field model to analyze the probability distribution characteristics of load demand and energy output, identify sensitive periods of high load fluctuations, low energy output, and extreme changes in environmental variables, and determine their impact on redundant capacity configuration by calculating the joint probability density of the variables within each time period; S2.2.1 In order to derive the redundant capacity thresholds for each time period, a redundant capacity demand model is constructed. By comprehensively analyzing the joint probability distribution of multiple variables, the optimal configuration range of redundant capacity, i.e., the redundant capacity demand, is determined. ; Specifically, the redundant capacity requirements At the moment Expressed as: ,in: Indicates the Power consumption nodes at time The load demand change; Indicates the Distributed energy at the moment Output change; Indicates the Environment variables at time The amount of change; is the total number of electricity consumption nodes, distributed energy resources and environmental variables; S2.2.2 Based on the redundant capacity demand model, derive the redundant capacity threshold for each time period ; Use a multi-objective optimization framework to consider power supply reliability, economy and resource utilization efficiency to construct an optimization objective function ,in: represents the power supply reliability function; represents the economic function; represents the resource utilization efficiency function; , , is the weight parameter, reflecting the relative importance of each goal; S2.2.3 During the optimization process, multi-level constraints and nonlinear optimization strategies are used to ensure the rationality and practicality of the redundant capacity threshold. The specific calculation formula is as follows: ,in: Indicates the Power consumption nodes at time the probability that the load demand does not exceed the redundancy threshold; is the cost function of redundant capacity, which is a nonlinear function reflecting the economic cost of redundant capacity; Indicates resource utilization efficiency and encourages efficient allocation of redundant capacity; S2.3.1 Based on the derived redundant capacity threshold, construct a time-sensitive redundant configuration strategy matrix , whose elements Indicates the time period within, no. The configuration amount of class redundant resources is specifically expressed as: ,in: is the total number of time periods divided, is the number of categories of redundant resources; S2.3.2 uses a method that combines hierarchical clustering and nonlinear optimization to generate a time-sensitive redundancy configuration strategy. The specific steps are as follows: Hierarchical clustering: Based on the redundant capacity threshold and load demand fluctuation characteristics within each time period, a hierarchical clustering algorithm is used to divide the time period into several clusters. The time periods within each cluster have similar redundant capacity demand characteristics. Non-linear optimization: For each cluster, a non-linear optimization algorithm is used to optimize the configuration ratio of redundant resources to ensure that the redundant capacity requirements of the cluster are met at all times while maximizing resource utilization efficiency and economy. Strategy mapping: Maps the optimization results to the redundancy configuration strategy matrix, providing specific values ​​for the redundant resource configuration within each time period, ensuring that the redundancy configuration strategy is targeted and sensitive in different time periods. S3. Based on the redundant capacity threshold, a Bayesian network is constructed to infer the energy node status, identify the synergistic relationship between various energy sources, and generate a dynamic redundant resource allocation plan to ensure the rational allocation of resources in multi-energy scenarios.

2. A method for using integrated services for intelligent design of electric power engineering according to claim 1, characterized in that: Step S1 includes the following contents: S1.1 First, through multi-source data collection and preprocessing, obtain real-time and historical data of various key parameters during microgrid operation; S1.2 In the time series dynamic characteristics modeling phase, the load demand changes are divided into several states, and the load demand transition probability matrix is ​​established using the Markov chain; The distributed energy output is modeled using an autoregressive hidden Markov model, which defines the hidden states and the corresponding output distribution. The model parameters are trained using the maximum expectation algorithm to characterize the intermittent and random nature of energy output. The spatiotemporal modeling of environmental variables adopts the high Vickrickin interpolation method, which describes the correlation of environmental variables between different geographical locations by defining the spatial covariance matrix and combines it with the time series analysis method to achieve dynamic adjustment of environmental variables.

3. The method for using the integrated service of intelligent design of electric power engineering according to claim 2, characterized in that: Step S1 also includes the following: S1.3 In the process of constructing the joint probability distribution, a multi-layer Copula function architecture is used to capture the nonlinear dependencies between load demand, distributed energy output, and environmental variables. First, a marginal distribution function is defined for each type of data. 、 、 , corresponding to load demand, energy output and environmental variables respectively; then, a two-layer Copula structure is constructed. The first layer adopts a hybrid form of Gaussian Copula and Clayton Copula to capture the tail dependence and symmetry between different variables respectively; the second layer refines the dynamic dependence relationship in the time series through conditional Copula; by introducing the time lag parameter , the joint probability distribution can reflect the current moment and the past Moment-to-moment dependencies; S1.4 In order to generate the joint probability distribution of multiple scenarios, the Markov chain Monte Carlo method is used to sample the constructed dynamic joint Copula model to generate a multi-scenario dataset covering different time periods and operating states; The multi-scenario dataset generated by S1.5 was statistically tested to verify the accuracy and generalization ability of the model, ensuring that the generated multi-scenario joint probability distribution can truly reflect the dynamic relationship of multiple variables in microgrid operation.

4. A method for using integrated services for intelligent design of electric power engineering according to claim 3, characterized in that: Step S3 includes the following contents: S3.1 First, identify and define various redundant resource nodes in the microgrid, including backup generators, energy storage equipment, renewable energy devices, and critical load nodes; S3.2 Based on the redundant resource node modeling, a Bayesian network is constructed to describe the conditional dependency relationship between nodes; S3.3 reflects the nonlinear dependency between redundant resource nodes through a conditional probability table; After completing the construction of the Bayesian network and its conditional probability table, S3.4 uses Bayesian reasoning technology and combines it with the redundant capacity threshold to optimize the allocation of dynamic redundant resources. The specific steps are as follows: Observation data input and posterior probability calculation: the load demand at the current moment , energy output and environment variables The observation data is input into the Bayesian network, and the posterior probability of each redundant resource node in different states is calculated by Bayesian theorem. ,in represents a set of observation data; Energy synergy factor calculation: Energy synergy factor , reflecting the synergistic effect between different energy types. The specific calculation formula is as follows: ,in: 、 Energy type and The output change at the current moment; is the attenuation parameter of the synergistic effect, which controls the attenuation rate of the synergistic effect with the output change difference; Energy type The actual output power; Energy type Maximum output power; S3.5 builds a multi-energy collaborative redundant resource allocation model with the goal of maximizing the overall power supply reliability and resource utilization efficiency of the system. The specific optimization objective function is defined as follows: ,in: is the number of redundant resource categories; 、 Redundant resource category and Configuration ratio; is the energy synergistic factor; 、 For nodes and The posterior probability of being in the running state.

5. The method for using the integrated service of intelligent design of electric power engineering according to claim 4, characterized in that: Step S3 also includes the following: S3.6 Under the defined objective function and constraints, solve the redundant resource allocation ratio The optimal value of , the specific optimization process includes: Initialization: Generate the initial population or starting point; Fitness evaluation: Calculate the fitness value of each configuration scheme, that is, the objective function The value of Selection and crossover: Select excellent individuals based on fitness, perform crossover and mutation operations, and generate a new generation of configuration solutions; Iterative optimization: Repeat the fitness evaluation and selection crossover process until the preset convergence condition or number of iterations is reached; Optimal solution extraction: The configuration solution with the highest fitness is finally selected as the optimal redundant resource allocation solution; S3.7 generates a specific redundant resource allocation plan based on the optimization results, and organizes the optimal configuration ratios into a redundant resource configuration matrix, where each element represents the configuration ratio of the redundant resource category at the current moment.

6. A power engineering intelligent design integrated service system, used to implement the power engineering intelligent design integrated service method according to any one of claims 1 to 5, characterized in that: include: Probabilistic modeling module, scenario generation module, sensitive identification module, threshold calculation module, network reasoning module and resource allocation module; Probabilistic modeling module: Constructs a dynamic joint probability field model of load demand, distributed energy output, and environmental variables, identifies the correlation and dynamic characteristics between variables, and generates a dynamic joint probability field model as the basis for the scenario generation module; Scenario Generation Module: Generates multi-scenario joint probability distribution from the probability model using Markov chain and Copula function; the output multi-scenario joint probability distribution is used for analysis by the Sensitive Recognition Module; Sensitive identification module: Based on joint probability field model analysis, it deduces sensitive periods of load fluctuation and energy output; The identification result is input into the threshold calculation module to determine the redundant capacity threshold; Threshold calculation module: uses dynamic programming algorithm to calculate the redundant capacity thresholds in different time periods; the output of the threshold calculation is sent to the network reasoning module for network reasoning and inferring the energy node status; Network reasoning module: Constructs a Bayesian network to reason about the status of energy nodes and identify the synergistic relationships among energy sources. The network reasoning results are used by the resource allocation module to generate dynamic allocation plans. Resource allocation module: Generates dynamic redundant resource allocation plans based on redundant capacity thresholds and inference results to ensure reasonable allocation of resources in multi-energy scenarios.

Images