Capacity expansion planning method and device applied to urban power grid, and medium

Abstract

The invention discloses a capacity expansion planning method and device applied to an urban power grid and a medium, and belongs to the technical field of urban power grid planning, and the method comprises the steps: extracting a typical scene through a continuous high and low temperature scene model which is established in advance according to an adequacy index; simulating the typical scene through a Markov chain and Monte Carlo method, and generating an annual typical scene sequence considering wind-light-water-load space-time correlation; inputting the yearly typical scene sequence into a capacity configuration model which is constructed in advance by taking the lowest yearly comprehensive cost of the system as a target to obtain an optimized capacity configuration model; the optimized capacity configuration model is converted into a mixed integer linear programming model, an optimal configuration scheme is solved through a commercial solver, and capacity expansion planning is completed through the optimal configuration scheme; the method can reduce the wind curtailment light quantity in a transition scene, increases the system adequacy under continuous high and low temperatures, and can make full use of the energy supply resources in the system to ensure the supply and demand balance of the power grid.

Description

Technical Field

[0001] The present invention relates to a capacity expansion planning method, device and medium applied to urban power grids, and belongs to the technical field of urban power grid planning. Background Art

[0002] With the acceleration of China's urbanization process, urban power grids have become a key link to ensure the electricity consumption of the vast number of residents. As the power load center, urban power grids are characterized by dense loads, high requirements for safety, reliability and power supply quality. However, under the background of the "dual carbon" goal drive and energy structure transformation, the large-scale access of distributed renewable energy shows strong time intermittency and power volatility. At present, the problems of insufficient flexibility and stability caused by planning defects in urban power grids are prominent. Under the background of the further intensification of the high proportion of distributed renewable energy penetration and the spatio-temporal imbalance of load distribution in the future, the problems of renewable energy consumption, load power supply guarantee and insufficiency in local time and local areas of urban power grids will be more serious.

[0003] In urban power grids with a high proportion of renewable energy, the output capacities of wind power, photovoltaic power and hydropower are all strongly affected by weather conditions. In addition, under the background of global warming, the significant increase in extreme temperature events has led to a sharp increase in electricity loads. In recent years, extreme temperature events such as cold snaps and heatwaves have caused power rationing and power outages in many countries. In August 2020, California, USA, experienced severe high-temperature weather, resulting in a sharp increase in load, a decrease in renewable energy output, and a large-scale power rolling blackout accident; in February 2021, Texas, USA, encountered extremely cold weather, resulting in a sharp increase in heating load due to the low temperature, the shutdown of wind turbines due to blade icing, and the freezing of natural gas wellheads affecting the operation of gas turbines, and forced to perform load shedding operations many times to maintain system stability; in August 2022, Sichuan encountered extremely high-temperature and drought weather, resulting in extremely insufficient hydropower supply and a sharp increase in cooling load, leading to an imbalance between power supply and demand, and ultimately resulting in large-scale power rationing.

[0004] The adequacy of the power generation system refers to the ability of the power generation system to continuously supply the total power demand and total electrical energy of users considering the planned or unplanned outages of system components, and is an important indicator to describe the reliability of the system. Traditionally, "loss of load probability" and "expected energy not served" are often used to describe the adequacy of the system. This indicator can better characterize the system load shedding situation, but ignores the adequacy evaluation of the system when no load shedding occurs.

[0005] In the face of the problem of insufficient long-term or short-term power generation capacity adequacy caused by the influence of the external environment on the power grid, corresponding measures can be taken on the power generation, grid, and load sides. On the supply side, through the expansion planning of tie lines, the interaction of electricity between adjacent power grids is used to alleviate the short-term adequacy shortage. However, the line investment cost is relatively large, and there is also the problem of low utilization rate of some lines after they are built and put into use. On the grid side, energy storage devices with the characteristics of energy time-shifting are used as a supplement to power generation adequacy, which can alleviate the adequacy shortage problem caused by the significant decline in the credible capacity of the system under the high penetration of renewable energy. On the load side, flexible loads participate in demand response under market incentives to reduce load demand and improve system adequacy. However, a large number of non-flexible loads do not participate in demand response, and the improvement effect of the overall system adequacy is limited. Summary of the Invention

[0006] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a capacity expansion planning method, device, and medium applied to urban power grids, which can reduce the amount of abandoned wind and light in transitional scenarios, increase the system adequacy under continuous high and low temperatures, and make full use of the energy supply resources in the system to ensure the balance between power supply and demand in the power grid.

[0007] To achieve the above object, the present invention is implemented by the following technical solutions:

[0008] In the first aspect, the present invention provides a capacity expansion planning method applied to urban power grids, including:

[0009] Identifying and extracting historical continuous high and low temperature scenarios and transitional scenarios through a continuous high and low temperature scenario model established in advance according to adequacy indicators, and clustering them into typical scenarios;

[0010] Simulating the typical scenarios through the Markov chain and Monte Carlo methods to generate an annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water, and load;

[0011] Inputting the annual typical scenario sequence into a capacity configuration model constructed in advance with the goal of minimizing the annual comprehensive cost of the system to obtain an optimized capacity configuration model;

[0012] Converting the optimized capacity configuration model into a mixed-integer linear programming model, and solving the optimal configuration plan through a commercial solver, and completing the capacity expansion planning through the optimal configuration plan.

[0013] Furthermore, the adequacy indicators include power adequacy and energy adequacy, where:

[0014] The calculation formula for the power adequacy is as follows:

[0015] ;

[0016] ;

[0017] Among them, is the power adequacy for a time period , and being 0 indicates that the system has no reserve. It is set that when the system has power adequacy; is the number of power sources put into operation; is the maximum output capacity of power source k in the time period; is the maximum output of energy storage in the time period; is the load power demand in the time period; is the maximum available power of the external interconnection system in the time period;

[0018] The calculation formula for the above-mentioned power adequacy is as follows:

[0019] ;

[0020] In the formula, is the power adequacy of the system within the evaluation period c, , and being 0 indicates that the system has no capacity surplus. It is set that when the system has power adequacy, is the total number of all operable power sources of the system; is the supply power of the external interconnection system in the time period;

[0021] When the power adequacy and energy adequacy within the evaluation period c both meet , at the same time, then the system is in an ideal adequacy state .

[0022] Furthermore, the method for establishing the continuous high and low temperature scenario model includes:

[0023] Dividing the temperature into a low temperature range , a high temperature range and a normal temperature range , and determining high and low temperature days according to the daily maximum / minimum temperature;

[0024] If the power adequacy and energy adequacy of the power grid are both lower than the average value of normal temperature days for several consecutive days, it is defined as a continuous high and low temperature scenario, as shown in the following formula:

[0025] ;

[0026] In the formula, is the average of the daily maximum or minimum temperature; is the number of days with continuous maximum or minimum temperature; is the starting date during the period of continuous high temperature or continuous low temperature; is the daily maximum or minimum temperature; is the grid adequacy status on the d-th day; is the set of grid adequacy statuses, indicating the set of grid adequacy statuses excluding ; is to satisfy and for the grid adequacy status interval, is the power adequacy status on the d-th day, is the electricity adequacy status on the d-th day; is the average of the daily power adequacy of the grid at normal temperature, is the average of the daily electricity adequacy of the grid at normal temperature.

[0027] Furthermore, the method for extracting the typical scenarios includes:

[0028] Step 1: Eliminate the sample days with large power fluctuations caused by faults or emergencies, and segment the original samples to obtain a set of grid fault-free historical data sequences for a certain season ;

[0029] Step 2: Set both the width of the observation window and its sliding step size to 1 day, and use the observation window to traverse all the historical sequences in, and screen the normal temperature day samples that satisfy in the observation window, and calculate the average values of the power and electricity adequacy of all normal temperature day samples , ;

[0030] Step 3: Extract the i-th sequence from , and identify its length as days; Let the position of the observation window , the number of days with continuous high or low temperature ;

[0031] Step 4: Identify whether the d-th day simultaneously satisfies , , , . If it is satisfied, let , slide the window, and let , and go to Step 7; if it is not satisfied, go to Step 5;

[0032] Step 5: If , mark the sequence of the consecutive days before the d-th day as a continuous high temperature scenario, and go to step 6; otherwise, directly go to step 6;

[0033] Step 6: Let , slide the window, and let ;

[0034] Step 7: If , then go to step 4; if , then go to step 5; otherwise, go to step 8;

[0035] Step 8: Reset the window position , modify the judgment condition , traverse the sequence again through steps 4 to 7 , and mark out the continuous low temperature scenario;

[0036] Step 9: Mark the remaining sample days in the sequence as transition scenarios;

[0037] Step 10: Set all transition scenarios as one class, and classify the continuous high temperature scenarios according to the number of consecutive days and the average temperature of the continuous high temperature scenarios , and identify the set of continuous high temperature scenario classes according to the temperature interval division. Similarly, obtain the set of continuous low temperature scenario classes;

[0038] Step 11: Traverse all elements in the set , and count the number of occurrences of each continuous high / low temperature scenario class and transition scenario class and the number of transitions between them;

[0039] Step 12: Through the improved k-means method, cluster the renewable energy output and load demand for each continuous high / low temperature scenario class and transition scenario class in turn to obtain the typical scenarios of the continuous high / low temperature scenario classes and transition scenario classes.

[0040] Further, the improved k-means method includes:

[0041] Step 1: Input the different scenario sets to be clustered;

[0042] Step 2: Input the number of cluster centers K for each type of scenario, and randomly select K samples as the initial cluster centers;

[0043] Step 3: Calculate the Euclidean distance between each sample and each cluster center, and classify each sample into the cluster with the closest distance according to the data of each sample;

[0044] Step 4: Calculate the distortion function:

[0045] ;

[0046] In the formula, is the data of sample i belonging to the k-th center; is the k-th clustering center; m is the number of samples; K is the number of clustering centers;

[0047] Step 5: Judge the change value of the distortion function after iteration whether it is less than the threshold

[0048] If so, jump to Step 7; ; Jump to Step 3 for the next iteration calculation; in the formula, is the number of samples belonging to the k-th center;

[0049] Step 7: Calculate the clustering validity index according to the clustering result;

[0050] ;

[0051] In the formula, is the k-th cluster; the numerator represents the average intra-cluster sample distance, and the denominator represents the minimum inter-cluster distance; the smaller the intra-cluster distance and the larger the inter-cluster distance, the better the clustering effect; therefore the smaller the index, the better the classification. By changing the number of clustering centers, using the index as the criterion, find the optimal number of clusters;

[0052] Step 8: Judge whether the value of K reaches the upper limit. If not, update the value of K and go to Step 2;

[0053] Step 9: Select the K value with the smallest value and its clustering result as the set of such typical scenarios;

[0054] Step 10: Judge whether all category scenarios have been clustered. If not, go to Step 1 until all category scenarios have been clustered.

[0055] Furthermore, the typical scenarios are simulated by the Markov chain and Monte Carlo method to generate an annual typical scenario sequence considering the spatio-temporal correlation of wind-light-water-load, including:

[0056] Step 1: Let the simulated season s = 1, the length of the scenario sequence n = 0, and identify the length of season s as days;

[0057] Step 2: Let t = 1, randomly select its typical clustering curve from the scenario class with the most occurrences within the season as the initial scenario state , and identify its length , let ;

[0058] Step 3: Let \(t = t + 1\), and search in the conditional probability matrix of scenario state transition for the set of state probabilities of transferring to the \(t\)-th scenario class. Extract a uniform distribution . If , then randomly extract a typical clustering curve from the scenario class and identify its length as ; ;

[0059] Among them, the single-step transfer conditional probability between scenario classes within season \(s\) and the conditional probability matrix of scenario state transition are calculated as follows:

[0060] ;

[0061] ;

[0062] Among them, is the total number of scenario classes included in season \(s\), is the number of occurrences of scenario class , is the probability of scenario class transferring to scenario class , is the number of times scenario class transfers to scenario class ;

[0063] Step 4: If , let , , and go back to Step 3; if , let \(t = t - 1\) and re-execute Step 3; if , let , output the current season's scenario sequence , and go to Step 5;

[0064] Step 5: If \(s \lt 4\), let \(s = s + 1\) and go back to Step 2; if \(s = 4\), output the annual typical scenario sequence .

[0065] Furthermore, the construction method of the capacity configuration model includes:

[0066] Taking the minimum of the system annual comprehensive cost composed of equal annual value investment cost, maintenance cost, operation cost, carbon emission cost and adequacy deficiency penalty cost as the goal, considering planning constraints, power and energy balance constraints, hydrogen energy storage operation constraints, thermal power unit operation constraints, and gas turbine operation constraints, establish a capacity configuration model, and the formula is as follows:

[0067] ;

[0068] Wherein, and are respectively the equivalent annual investment cost and annual maintenance cost of the newly added power source and energy storage in the distribution network; and and are respectively the annual operation cost, annual carbon emission cost and annual penalty cost for insufficient adequacy of the configured distribution network.

[0069] Furthermore, the step of converting the optimized capacity configuration model into a mixed-integer linear programming model and solving the optimal configuration scheme through a commercial solver includes:

[0070] Introduce auxiliary variables and the big M method to convert the non-linear constraints of 0-1 variables and continuous variables into linear constraints;

[0071] Adopt the Benders decomposition method to iteratively solve the master problem and the sub-problem to approximate the global optimal solution and obtain the optimal configuration scheme.

[0072] In a second aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, characterized in that: when the program is executed by a processor, the steps of any one of the foregoing methods are implemented.

[0073] In a third aspect, the present invention provides a capacity expansion planning device applied to an urban power grid, including:

[0074] A memory for storing computer programs / instructions;

[0075] A processor for executing the computer programs / instructions to implement the steps of any one of the foregoing methods.

[0076] Compared with the prior art, the beneficial effects achieved by the present invention:

[0077] The present invention provides a capacity expansion planning method, device and medium for urban power grids. By simulating typical scenarios through Markov chains and Monte Carlo methods, an annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water and load is generated. The annual typical scenario sequence is input into a capacity configuration model previously constructed with the goal of minimizing the annual comprehensive cost of the system to obtain an optimized capacity configuration model. The optimized capacity configuration model is converted into a mixed-integer linear programming model, and the optimal configuration plan is solved through a commercial solver. The capacity expansion planning is completed through the optimal configuration plan. Through the present invention, the amount of abandoned wind and light in transitional scenarios can be reduced, the system adequacy under continuous high and low temperatures can be increased, the energy supply resources in the system can be fully utilized to ensure the balance between power supply and demand of the power grid, and it is significantly helpful for improving the system flexibility and the power supply and demand security of the system under abnormal environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 is a flowchart of a capacity expansion planning method for urban power grids provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0079] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and cannot be used to limit the protection scope of the present invention.

[0080] Embodiment 1. This embodiment introduces a capacity expansion planning method for urban power grids, including:

[0081] By using a continuous high and low temperature scenario model established in advance according to adequacy indicators, historical continuous high and low temperature scenarios and transitional scenarios are identified and extracted, and clustered into typical scenarios;

[0082] By simulating typical scenarios through Markov chains and Monte Carlo methods, an annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water and load is generated;

[0083] The annual typical scenario sequence is input into a capacity configuration model previously constructed with the goal of minimizing the annual comprehensive cost of the system to obtain an optimized capacity configuration model;

[0084] The optimized capacity configuration model is converted into a mixed-integer linear programming model, and the optimal configuration plan is solved through a commercial solver. The capacity expansion planning is completed through the optimal configuration plan.

[0085] As Figure 1 shown, the capacity expansion planning method for urban power grids provided in this embodiment specifically involves the following steps in its application process:

[0086] Establish adequacy indicators to model continuous high and low temperature scenarios;

[0087] Identify and extract historical continuous high and low temperature scenarios and transition scenarios through a continuous high and low temperature scenario model, and cluster them into typical scenarios;

[0088] Obtain the annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water, and load through the Markov Chain Monte Carlo method;

[0089] With the goal of minimizing the annual comprehensive cost of the system, considering planning constraints, power and energy balance constraints, and equipment operation constraints, establish an optimization configuration model for power supply and energy storage capacity based on annual 8760h panoramic time series simulation;

[0090] Convert the capacity optimization decision model considering the coordinated configuration of power supply and energy storage into a mixed-integer linear programming model and solve it using existing commercial solvers.

[0091] Specifically, establishing adequacy indicators to model continuous high and low temperature scenarios includes:

[0092] Evaluate the supply-demand state of the power grid by establishing adequacy indicators. Consider power and energy adequacy based on temperature to evaluate the impact of continuous high and low temperature weather on the power system, and then establish a mathematical model of continuous high and low temperature scenarios for urban power grids.

[0093] Specifically, the adequacy indicators include:

[0094] Define the ratio of the reserve capacity of the system at each time period to the load as the power adequacy of the system at that time period, which is used to characterize the adequacy of the system's power supply capacity, as shown in Equation (1). The power adequacy of each time period will change with the fluctuations of new energy generation and load. Select the minimum value of the power adequacy of each time period in the evaluation period as the power adequacy of the entire period, as shown in Equation (2).

[0095] ;

[0096] ;

[0097] In the formula, is the power adequacy of time period , , and 0 means the system has no reserve. Referring to the load and fault reserve that should be reserved during the normal operation of the system, it is considered that when the system has power adequacy, where >0 can be taken as the typical heat reserve coefficient of the system; is the number of power supplies put into operation; is the maximum output capacity of power supply k in time period . For conventional power supplies, the rated output is taken, and for new energy power supplies, the maximum output under the wind and light resource conditions of that time period is taken; is The maximum output of energy storage during a period is the load power demand during a period is the maximum available power that can be supplied by the external interconnected system during a period is the system's power adequacy during the evaluation period c

[0098] The system's power adequacy is defined by the ratio of the system's reserve power consumption to the load demand power consumption during the evaluation period, and is used to characterize the adequacy of the power supply volume of the system during the evaluation period, as shown in Equation (3).

[0099] ;

[0100] In the formula, is the system's power adequacy during the evaluation period c , and when it is 0, it means that the system has no capacity surplus. Considering that a certain reserve capacity needs to be reserved for a normally operating system, it is set that when the system has power adequacy, where >0 can refer to the typical cold and hot reserve settings required by the system is the total number of all operable power sources in the system is the power supply of the external interconnected system during a period is the total storage capacity of the system's energy storage at the initial time of the evaluation period c

[0101] When the power and power adequacy during the evaluation period c simultaneously satisfy , then it is considered that the system is in an ideal adequacy state .

[0102] Specifically, the modeling of continuous high and low temperature scenarios includes:

[0103] According to the analysis of the impact of ambient temperature on the power sources and loads of the urban power grid, the low temperature range is divided into , , ; the high temperature range is divided into , , ; the normal temperature range is divided into . If the daily maximum (minimum) temperature is in the high temperature range (low temperature range), then the day is called a high temperature day (low temperature day), otherwise it is a normal temperature day

[0104] Set the evaluation period to one day, and use the average values of the daily power and power adequacy of the power grid at normal temperature , as a reference. If the power grid operates for multiple days and simultaneously meets the condition of continuously being in a high (low) temperature range, and the power grid adequacy state continuously does not reach the ideal state , 、 are respectively continuously less than 、 , then use the number of days of continuous high (low) temperature and the average value of the highest (lowest) temperature of each day during the continuous high (low) temperature period to describe the continuous high (low) temperature scenario, as shown in Equation (4).

[0105] ;

[0106] In the formula, is the starting date during the continuous high temperature or continuous low temperature period; is the daily highest or lowest temperature; is the power grid adequacy state on the d-th day; is the set of power grid adequacy states, represents the set of power grid adequacy states except ; is the power grid adequacy state interval that satisfies and .

[0107] Specifically, the method for extracting the characteristics of typical scenarios includes:

[0108] Based on the historical operation data of the power grid in each season, identify its typical operation scenarios, and obtain the typical power curves of wind power, photovoltaic power, hydropower, and load under the continuous high (low) temperature scenarios and transition scenarios of the power grid considering spatio-temporal correlation in each season. The steps for extracting the characteristics of typical scenarios in season s of the power grid are as follows:

[0109] Step 1: Eliminate the sample days containing large power fluctuations caused by faults or emergencies, and segment the original samples accordingly to obtain the set of historical data sequences of the fault-free power grid in a certain season .

[0110] Step 2: Set the width of the observation window and its sliding step size to 1 day each, and use the observation window to traverse all the historical sequences in, and screen the normal temperature day samples that satisfy in the observation window, and calculate the average values of power and electricity adequacy of all normal temperature day samples 、 .

[0111] Step 3: Extract the i-th sequence from , and identify its length as days. Let the position of the observation window Number of consecutive days of high and low temperatures .

[0112] Step 4: Identify whether the d-th day simultaneously satisfies , , , . If satisfied, let , slide the window, let , and go to Step 7; if not satisfied, go to Step 5.

[0113] Step 5: If , mark the sequence of the consecutive days before the d-th day as a continuous high-temperature scenario, and go to Step 6; otherwise, directly go to Step 6.

[0114] Step 6: Let , slide the window, let .

[0115] Step 7: If , then go to Step 4; if , then go to Step 5; otherwise, go to Step 8.

[0116] Step 8: Reset the window position , modify the judgment condition , traverse the sequence again through Steps 4 to 7 , and mark out the continuous low-temperature scenario.

[0117] Step 9: Mark the remaining sample days in the sequence as transition scenarios.

[0118] Step 10: Set all transition scenarios as one class, and classify the continuous high-temperature scenarios according to the number of consecutive days and average temperature of the continuous high-temperature scenarios , and identify the set of continuous high-temperature scenario classes according to the temperature interval division in Section 1.3. Similarly, obtain the set of continuous low-temperature scenario classes.

[0119] Step 11: Traverse all elements in the set , and count the number of occurrences of each continuous high (low)-temperature scenario class and transition scenario class, as well as the number of transitions between them.

[0120] Step 12: Through the improved k-means method, perform clustering of renewable energy output and load demand for each continuous high (low)-temperature scenario class and transition scenario class in turn, and obtain the typical clustering curves of the continuous high (low)-temperature scenario classes and transition scenario classes.

[0121] Specifically, the improved k-means method includes:

[0122] Step 1: Input different scenario sets to be clustered.

[0123] Step 2: Input the number K of clustering centers for each type of scenario, and randomly select K samples as the initial clustering centers.

[0124] Step 3: Calculate the Euclidean distance between each sample and each clustering center, and classify each sample into the cluster with the nearest distance according to the data of each sample.

[0125] Step 4: Calculate the distortion function:

[0126] ;

[0127] In the formula, is the data of sample i belonging to the k-th center; is the k-th clustering center; m is the number of samples; K is the number of clustering centers.

[0128] Step 5: Judge whether the change value of the distortion function after iteration is less than the threshold . If so, jump to Step 7.

[0129] Step 6: Update the clustering center as shown in Equation (6), jump to Step 3, and perform the next iteration calculation.

[0130] ;

[0131] In the formula, is the number of samples belonging to the k-th center.

[0132] Step 7: According to the clustering result, calculate the clustering validity index .

[0133] ;

[0134] In the formula, is the k-th cluster; the numerator represents the average distance between samples within the cluster, and the denominator represents the minimum distance between clusters. The smaller the distance within the cluster and the larger the distance between clusters, the better the clustering effect. Therefore, the smaller the index, the better the classification. By changing the number of clustering centers, using the index as the criterion, find the optimal number of clusters. Generally, the value range of the number of clusters K is .

[0135] Step 8: Judge whether the value of K reaches the upper limit. If not, update the value of K and go to Step 2.

[0136] Step 9: Select The K values with the smallest values and their clustering results are used as the set of such typical scenarios.

[0137] Step 10: Determine whether all category scenarios have been clustered. If not, go to Step 1.

[0138] Specifically, the method for generating the annual typical scenario sequence includes:

[0139] Let the total number of scenario categories in season s be , and the number of times scenario category i appears be . The probability and number of times scenario category i transfers to scenario category j are and respectively. Then the single-step transfer conditional probability between scenario categories within season s and the scenario state transition conditional probability matrix can be expressed as:

[0140] ;

[0141] ;

[0142] The Markov Chain Monte Carlo method is used to generate the future annual typical scenario sequence. The specific steps are as follows:

[0143] Step 1: Let the simulated season s = 1 and the scenario sequence length n = 0. Identify the length of season s as days.

[0144] Step 2: Let t = 1. Randomly select its typical clustering curve from the scenario category with the most occurrences within the season as the initial scenario state , and identify its length . Let .

[0145] Step 3: Let t = t + 1. In the scenario state transition conditional probability matrix , find the set of state probabilities for the state transferring to the t-th scenario category. Extract a uniform distribution . If , then randomly select a typical clustering curve from the scenario category , and identify its length as .

[0146] Step 4: If , let , , and go to Step 3; if , let t = t - 1, and re-execute Step 3; if , let , output the current seasonal scene sequence , go to step 5.

[0147] Step 5: If s < 4, s = s + 1, go to step 2; if s = 4, output the annual typical scene sequence .

[0148] Specifically, the expansion planning method for power supply and energy storage includes:

[0149] With the goal of minimizing the total annual comprehensive cost of the system composed of equal annual value investment cost, maintenance cost, operation cost, carbon emission cost and penalty cost for insufficient adequacy, considering planning constraints, power and electricity balance constraints, hydrogen energy storage operation constraints, thermal power unit operation constraints, and gas turbine operation constraints, a capacity optimization configuration model for power supply and energy storage based on annual 8760h panoramic time series simulation is established.

[0150] Specifically, the expansion planning goals for distribution network power supply and energy storage include:

[0151] With the goal of minimizing the sum of the equal annual value planning cost of the power supply and energy storage capacity configuration over the entire life cycle and the total annual operation cost of the system after configuration , a capacity expansion planning model for power supply and energy storage is established as shown in Equation (10).

[0152] ;

[0153] In the formula, , are respectively the equal annual value investment cost and annual maintenance cost of the newly added power supply and energy storage in the distribution network; , , are respectively the annual operation cost, annual carbon emission cost and annual penalty cost for insufficient adequacy of the distribution network after configuration.

[0154] (1) Equal annual value investment cost

[0155] ;

[0156] In the formula, r is the discount rate; is the service life of equipment k; is the configured capacity of equipment k; is the unit capacity configuration cost coefficient of equipment k; k takes 1 to 5 respectively representing the related equipment EL, HST, FC, GT and TP of the power supply and energy storage that may be newly added to the distribution network.

[0157] (2) Maintenance cost

[0158] ;

[0159] In the formula, is the annual maintenance cost per unit power or capacity of equipment k.

[0160] (3) Operating cost

[0161] The system operating cost consists of curtailment cost, fuel purchase cost, and load shedding cost, as shown in Eqs. (13)-(14).

[0162] ;

[0163] ;

[0164] In the formula, , , are the curtailment cost, fuel purchase cost, and load shedding cost of the distribution network at time t, respectively; , , , are the curtailment power, GT power generation, EL power, and load shedding amount of the distribution network at time t, respectively; , are the curtailment penalty unit price and load shedding penalty unit price, respectively; , are the unit fuel consumption prices of GT and thermal power per unit of electricity, respectively; is the water consumption cost per unit of electricity for EL hydrogen production; is the unit time length.

[0165] (4) Carbon emission cost

[0166] The total carbon emissions considered in this embodiment consists of the carbon emissions generated by thermal power units and gas turbine power generation, as shown in Eq. (15).

[0167] ;

[0168] In the formula, , are the carbon emission coefficients of thermal power units and gas turbine power generation, respectively.

[0169] The carbon emission cost is the carbon emission fee paid proportionally for the part where the total carbon emissions of the system power generation exceed its free carbon quota, as shown in Eq. (16).

[0170] ;

[0171] In the formula, is the carbon trading price, with the unit of (yuan / ton); is the free carbon emission quota of the system.

[0172] (5) Insufficiency penalty cost

[0173] In this embodiment, with a day as the observation period, the power and electricity insufficiency penalty costs are calculated respectively, as shown in formulas (17)-(19).

[0174] ;

[0175] ;

[0176] ;

[0177] In the formulas, , are respectively the power insufficiency penalty cost and the electricity insufficiency penalty cost of the distribution network on the d-th day; is the power insufficiency penalty coefficient; is the electricity insufficiency penalty coefficient; , are respectively the power and electricity sufficiency degrees of the distribution network on the d-th day.

[0178] Specifically, the expansion planning constraints include:

[0179] The equivalent annual value planning cost of various devices in the distribution network needs to meet a certain upper limit, as shown in formula (20). The total planned power of the gas turbine should be restricted by the gas network carrying capacity, and the configuration of thermal power units and hydrogen energy storage should meet the upper limit of capacity configuration, as shown in formulas (21), (22), and (23) respectively.

[0180] ;

[0181] ;

[0182] ;

[0183] ;

[0184] In the formulas, is the maximum equivalent annual investment budget; M is the scheduling period, which is set to 8760h in this embodiment; is The maximum output capacity of GT during a period is limited by the real-time gas supply volume of the gas network; It is the upper limit of the HST capacity configuration; 、 、 They are the power configuration upper limits of TP, EL, and FC respectively.

[0185] Specifically, the power and electricity balance constraints include:

[0186] Based on the typical power curve of 8760h throughout the year, an hourly power balance constraint is established:

[0187] ;

[0188] ;

[0189] In the formula, is the distributed wind and solar power fed into the grid during the t period; is the FC output during the period; is the total power generation of thermal power units during the t period; 、 are the operating states of EL and FC during the t period respectively.

[0190] Considering the long-term energy storage characteristics of the hydrogen energy storage, a monthly electricity balance constraint for the urban power grid is constructed:

[0191] ;

[0192] In the formula, is the electricity demand of the distribution network in the m-th month; is the electricity consumption for hydrogen production by EL in the m-th month; is the load shedding amount in the m-th month; is the increase in hydrogen storage in the m-th month; is the electricity conversion loss from electricity to hydrogen and then back to electricity; is the hydrogen production efficiency of water electrolysis by EL; is the FC power generation in the m-th month; is the electricity transmitted from the transmission grid in the m-th month; is the distributed wind and solar power fed into the distribution network in the m-th month; is the thermal power generation in the m-th month; is the GT power generation in the m-th month; is the electricity wasted by the power grid in the m-th month.

[0193] Specifically, the operating constraints of the hydrogen energy storage include:

[0194] The storage state of hydrogen energy storage during operation needs to satisfy the constraints shown in Equations (27)-(30).

[0195] ;

[0196] ;

[0197] ;

[0198] ;

[0199] In the formula, is the change in the state of hydrogen (SOH) storage in the HST at time t; is the configured hydrogen storage capacity (tons) of the HST; and are the hydrogen storage and release efficiencies respectively; and are the initial and final hydrogen storage states of the HST at the start of an entire dispatching cycle respectively.

[0200] The net load of the urban power grid fluctuates frequently and has high requirements for power supply quality. Proton exchange membrane electrolyzers and alkaline fuel cells with flexible regulation rate and high maturity are used as hydrogen-electric conversion devices, and their conversion constraints are shown in Equations (31) and (32).

[0201] ;

[0202] ;

[0203] In the formula, is the hydrogen production amount of the EL at time t; is the hydrogen consumption amount of the FC at time t; is the power generation efficiency of the FC.

[0204] Both the electrolyzer and the fuel cell need to satisfy the upper and lower limits of electric power shown in Equation (33); the electric power ramp constraint shown in Equation (34); the start-stop times constraint shown in Equation (35); the minimum continuous operation and shutdown time constraint shown in Equation (36); and the hydrogen charging and discharging rate constraint shown in Equation (37).

[0205] ;

[0206] ;

[0207] ;

[0208] ;

[0209] ;

[0210] wherein, is the rated power configured for the device ; is the operating state of the device at time period t, and the value of 1 indicates being in the operating state, and 0 indicates being in the shutdown state; is the minimum load rate of the device ; is the upper limit of the device ramp per unit time; is the number of time periods in a day; and are the start and stop action variables of the device ; when is 1, the device starts at the start time period of time period t, and when is 1, the device shuts down at time period t; is the daily start-stop times limit of the device ; , are the continuous operation and shutdown times of the device at time period t respectively; , are the minimum operation and shutdown times of the device respectively; is the magnification coefficient of the HST configuration amount and the hydrogen charging and discharging speed.

[0211] In addition, to prevent the fuel cell from experiencing hydrogen starvation, the lower limit of the hydrogen storage amount during the operation of the fuel cell is restricted, as shown in Equation (38).

[0212] ;

[0213] wherein, is the lower limit of the hydrogen storage amount for maintaining the operation of the FC.

[0214] Specifically, the power operation constraints include:

[0215] ​​The gas turbine and the thermal power unit need to satisfy the upper and lower limits of output constraints shown in Equations (39) and (40), respectively.

[0216] ;

[0217] ;

[0218] In the formula, and are the minimum operating powers of GT and TP, respectively.

[0219] In addition, both the gas turbine and the thermal power unit need to satisfy the ramping constraints and start-stop constraints shown in Equations (41)-(42).

[0220] ;

[0221] ;

[0222] In the formula, is the output of the equipment in the time period; and are the upper and lower limits of the ramping power of the equipment per unit time, respectively, for the equipment.

[0223] Specifically, the model solution method includes:

[0224] In this embodiment, the planning based on the annual typical scenario sequence simulates and optimizes the system time-series operation state of the planning horizon year with hourly resolution while optimizing the capacity configuration, resulting in a large computational burden. The Benders decomposition method is adopted, where the master problem represents the equipment capacity optimization problem and the sub-problem represents the operation optimization problem under known capacity conditions. Since there are many variables and constraints in the 8760h optimization sub-problem, the operation optimization is divided into multiple sub-problems month by month. The cutting planes obtained by dual-solving each sub-problem are added to the master problem to improve the lower bound of the master problem, and the master problem and the sub-problem are iterated multiple times to approximate the global optimal solution.

[0225] In addition, to maintain the efficiency and accuracy of the solution, the planning model needs to be linearized. For the non-linear constraint caused by the multiplication of the 0-1 variable and the continuous variable in the constraint, a new variable is introduced to replace , and after linearization using the big M method, it is shown in Equation (43).

[0226] ;

[0227] In this embodiment, for all non-linear constraints involving the multiplication of 0-1 variables and continuous variables, the above method is used for linearization.

[0228] In the face of the power grid supply-demand changes under continuous high and low temperature environments, this embodiment proposes a capacity expansion planning method and system that takes into account the impact of continuous high and low temperatures on the power supply adequacy of urban power grids. Power and electricity adequacy indicators are established to evaluate the supply-demand relationship during the observation period of the system, and a continuous high and low temperature scenario model for urban power grids is further established from the perspective of adequacy; through typical scenario clustering and Markov Monte Carlo simulation, an annual typical scenario sequence considering the spatio-temporal correlation of wind-solar-hydro-load and including continuous high and low temperature scenarios is obtained; based on the obtained annual typical scenario sequence, with the goal of minimizing the annual comprehensive cost of the system, considering planning constraints, power and electricity balance constraints, and equipment operation constraints, a power source and energy storage capacity optimization configuration model based on annual 8760h panoramic time series simulation is established; the configuration model is converted into a mixed-integer linear programming model and solved using existing commercial solvers. Through the expansion planning decision method of this patent invention, the curtailment of wind and solar power in transitional scenarios can ultimately be reduced, the system adequacy under continuous high and low temperatures can be increased, and the energy supply resources in the system can be fully utilized to ensure the power grid supply-demand balance. In summary, a capacity expansion planning method and system that takes into account the impact of continuous high and low temperatures on the power supply adequacy of urban power grids provided by the present invention is of significant help in improving the system flexibility and system supply-demand security under abnormal environments.

[0229] Embodiment 2 provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the method described in any one of Embodiment 1 are implemented.

[0230] Embodiment 3 provides a capacity expansion planning device applied to urban power grids, including:

[0231] A memory for storing computer programs / instructions;

[0232] A processor for executing the computer programs / instructions to implement the steps of the method described in any one of Embodiment 1.

[0233] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

[0234] Those skilled in the art should understand that the embodiments of the present disclosure may be provided as a method, a system, or a computer program product. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0235] The present disclosure is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0236] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0237] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0238] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit the scope of its protection. Although the present disclosure has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that after reading the present disclosure, various changes, modifications, or equivalent replacements can still be made to the specific implementation manners of the invention. However, these changes, modifications, or equivalent replacements are all within the scope of the protection of the pending claims of the disclosure.

Claims

1. A capacity expansion planning method applied to urban power grids, characterized in that, Including: Identifying and extracting historical continuous high and low temperature scenarios and transition scenarios through a continuous high and low temperature scenario model established in advance according to adequacy indicators, and clustering them into typical scenarios; Simulating the typical scenarios through Markov chain and Monte Carlo methods to generate an annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water, and load; Inputting the annual typical scenario sequence into a capacity configuration model constructed in advance with the goal of minimizing the annual comprehensive cost of the system to obtain an optimized capacity configuration model; Converting the optimized capacity configuration model into a mixed-integer linear programming model, and solving the optimal configuration plan through a commercial solver, and completing the capacity expansion plan through the optimal configuration plan.

2. The capacity expansion planning method applied to urban power grids according to claim 1, wherein, The adequacy indicators include power adequacy and energy adequacy, where: The calculation formula of the power adequacy is as follows: ； ； Among them, is the power adequacy for the time period . When it is 0, it means the system has no reserve. It is set that the system has power adequacy; is the number of power sources put into operation; is the maximum output capacity of power source k during the time period ; is the maximum output of energy storage during the time period ; is the load power demand during the time period ; is the maximum available power of the external interconnection system during the time period ; is the system power adequacy within the evaluation period c; ​ The calculation formula of the energy adequacy is as follows: ； Wherein, is the system power adequacy during the evaluation period c, , 0 indicates that the system has no capacity surplus. It is set that when the system has power adequacy, is the total number of all operable power sources in the system; is the supply power of the external interconnected system during the period; is the total storage capacity of the system energy storage at the initial time of the evaluation period c; When the power and electricity adequacy within the evaluation period c simultaneously meet , , the system is in an ideal adequacy state .

3. The capacity expansion planning method applied to urban power grids according to claim 2, characterized in that The method for establishing the continuous high and low temperature scenario model includes: Divide the temperature into a low-temperature range , a high-temperature range and a normal-temperature range , and determine high / low-temperature days based on the daily maximum / minimum temperature; If the power adequacy and energy adequacy of the power grid are both lower than the average value of normal temperature days for multiple consecutive days, it is defined as a continuous high and low temperature scenario, as shown in the following formula: ； Wherein, is the average value of the daily maximum or minimum temperature; is the number of days of continuous maximum or minimum temperature; is the starting date during the period of continuous high temperature or continuous low temperature; is the daily maximum or minimum temperature; is the grid adequacy status on the d-th day; is the set of grid adequacy statuses, denotes the set of grid adequacy statuses excluding ; is the grid adequacy status interval that satisfies and ; is the power adequacy status on the d-th day, is the electricity adequacy status on the d-th day; is the average value of the daily power adequacy of the grid at normal temperature, is the average value of the daily electricity adequacy of the grid at normal temperature.

4. The capacity expansion planning method for urban power grids according to claim 3, wherein The method for extracting the typical scenarios includes: Step 1: Eliminate the sample days containing significant power fluctuations caused by faults or emergencies, and segment the original samples accordingly to obtain a set of historical fault-free power grid data sequences for a certain season. ; Step 2: Set the width of the observation window and its sliding step size to 1 day each, and use the observation window to traverse all historical sequences in , screen the normal temperature day samples that meet the requirements within the observation window, and calculate the average values of power and power adequacy of all normal temperature day samples , ; Step 3: Extract the i-th sequence from , identify its length as days; Let the observation window position be , the number of consecutive high and low temperature days be , ; Step 4: Identify whether the d-th day simultaneously satisfies , , , . If satisfied, let , slide the window, let , and go to Step 7; if not satisfied, go to Step 5. Step 5: If , mark the sequence of the consecutive days before the d-th day as a continuous high temperature scenario, and go to Step 6; otherwise, directly go to Step 6; Step 6: Let , slide the window, and let ; Step 7: If , then go to Step 4; if , then go to Step 5; otherwise, go to Step 8; Step 8: Reset the window position , modify the judgment condition , traverse the sequence again through Steps 4 to 7 , mark the continuous low-temperature scenarios; Step 9: Mark the remaining sample days in the sequence as transitional scenarios; Step 10: Classify all transition scenarios into one category. According to the number of consecutive days and average temperature of the continuous high-temperature scenario , classify the continuous high-temperature scenarios according to the temperature range division, and identify the set of continuous high-temperature scenario classes. Similarly, obtain the set of continuous low-temperature scenario classes; Step 11: Traverse all elements in the set , count the occurrences of each continuous high / low temperature scenario class and transition scenario class, as well as the number of transitions between them; Step 12: Through the improved k-means method, clustering the renewable energy output and load demand of each continuous high / low temperature scenario class and transition scenario class in turn to obtain the typical scenarios of the continuous high / low temperature scenario class and transition scenario class.

5. The capacity expansion planning method for urban power grid according to claim 4, characterized in that The improved k-means method includes: Step 1: Input different scenario sets to be clustered; Step 2: Input the number of cluster centers K for each type of scenario, and randomly select K samples as the initial cluster centers; Step 3: Calculate the Euclidean distance between each sample and each cluster center, and classify each sample into the cluster with the closest distance according to the data of each sample; Step 4: Calculate the distortion function: ； Wherein, is the data of sample i belonging to the k-th center; is the k-th clustering center; m is the number of samples; K is the number of clustering centers; Step 5: Determine the change value of the distortion function after iteration Is it less than the threshold If so, jump to Step 7; Step 6: Update the cluster centers as shown in the following formula: ; Jump to Step 3 for the next iterative calculation; In the formula, is the number of samples belonging to the k-th center; Step 7: Calculate the clustering validity index based on the clustering results ; ； In the formula, is the k-th cluster; the numerator represents the average distance between samples within the cluster, and the denominator represents the minimum distance between clusters; the smaller the distance within the cluster and the larger the distance between clusters, the better the clustering effect; thus the smaller the index, the better the classification. By changing the number of clustering centers, using the index as a criterion, the optimal number of clusters is found; Step 8: Judge whether the value of K reaches the upper limit. If not, update the value of K and go to Step 2; Step 9: Select the K value with the smallest value and its clustering result as the set of typical scenarios for this category; Step 10: Judge whether the clustering of all types of scenarios is completed. If not, go to Step 1 until the clustering of all types of scenarios is completed.

6. The capacity expansion planning method applied to urban power grids according to claim 5, characterized in that, The simulation of the typical scenarios through Markov chain and Monte Carlo methods to generate an annual typical scenario sequence considering the spatio-temporal correlation of wind, light, water, and load includes: Step 1: Set the simulated season s = 1 and the length of the scene sequence n = 0, and identify that the length of season s is days; Step 2: Let t = 1, randomly select the typical clustering curve from the scenario class with the most occurrences within the season as the initial scenario state , and identify its length , let ; Step 3: Let t = t + 1, and search in the scenario state transition conditional probability matrix for the set of state probabilities of transferring the state to the t-th scenario class, extract a uniform distribution , and if , then randomly extract a typical clustering curve from the scenario class and identify its length as ; Among them, the single-step transfer conditional probability between scene classes within season s and the scene state transition conditional probability matrix are calculated as follows: ； ； Among them, is the total number of scene classes included in season s, is the scene class the number of occurrences, is the scene class transfer to scene class probability, is the scene class transfer to scene class number of times; Step 4: If , let , , and go to Step 3; if , let t = t - 1, and re - execute Step 3; if , let , output the current seasonal scene sequence , and go to Step 5; Step 5: If s < 4, then s = s + 1, and go to Step 2; if s = 4, output the annual typical scenario sequence .

7. The capacity expansion planning method for urban power grids according to claim 6, characterized in that The method for constructing the capacity configuration model includes: With the goal of minimizing the annual comprehensive cost of the system composed of equal annual value investment cost, maintenance cost, operation cost, carbon emission cost, and adequacy deficiency penalty cost, considering planning constraints, power and energy balance constraints, hydrogen energy storage operation constraints, thermal power unit operation constraints, and gas turbine operation constraints, establish a capacity configuration model, the formula is as follows: ； Wherein, and are respectively the equal annual value investment cost and annual maintenance cost of the newly added power source and energy storage in the distribution network; and and are respectively the annual operating cost, annual carbon emission cost and annual adequacy deficiency penalty cost of the configured distribution network.

8. The capacity expansion planning method applied to urban power grids according to claim 7, characterized in that, The conversion of the optimized capacity configuration model into a mixed-integer linear programming model and the solution of the optimal configuration plan through a commercial solver include: Introducing auxiliary variables and the big M method to convert the non-linear constraints of 0-1 variables and continuous variables into linear constraints; Adopting the Benders decomposition method to iteratively solve the master problem and the sub-problem to approximate the global optimal solution and obtain the optimal configuration plan.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it realizes the steps of the method described in any one of claims 1-8.

10. A capacity expansion planning device applied to an urban power grid, characterized in that, Including: A memory for storing computer programs / instructions; A processor for executing the computer program / instructions to implement the steps of the method according to any one of claims 1-8.

Images