Peak regulation demand typical scene generation method and device for high-proportion new energy power system planning, and medium

Abstract

The invention discloses a peak regulation demand typical scene generation method and device for high-proportion new energy power system planning, and a medium, and belongs to the technical field of power system planning and operation, and the method comprises the steps: collecting time sequence data of a new energy power system, and carrying out the preprocessing; constructing an extreme value sequence to obtain a relative peak-valley difference matrix; clustering by taking the relative peak-valley difference matrix as a characteristic index to obtain a clustering result; extracting a new energy power system time sequence data curve of the representative day according to a clustering result; and combining the time sequence data curves of the new energy power system to obtain a net load representative day scene of the high peak regulation demand of the target year. The method has the beneficial effects that the definition of the relative peak-valley difference matrix is provided, the relative peak-valley difference matrix is used as a characteristic index of a clustering algorithm, wind / light / load historical sample data is grouped, and daily power curves with similar peak value and valley value appearing time periods and similar intra-day maximum peak-valley difference can be more effectively clustered into one class.

Description

Technical Field

[0001] This invention relates to the field of power system planning and operation technology, specifically to a method, system, equipment, and medium for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning. Background Technology

[0002] The high volatility, uncertainty, and unpredictability of renewable energy power generation pose significant challenges to peak-shaving and dispatching in new power systems. For planned high-proportion renewable energy power systems, assessing peak-shaving demand is a crucial step in determining the allocation of flexible adjustment resources such as energy storage.

[0003] Peak shaving corresponds to the balance of power supply and demand during the peak and valley periods of the grid's net load curve. The net load curve is calculated by algebraically superimposing the power generation of wind and solar power, which is considered to be unaffected by power generation plans, under the condition of maximizing the utilization of wind and solar resources, onto the load demand. The net load curve represents the power supply demand that needs to be adapted to by flexible power generation resources such as conventional power sources, pumped storage, and electrochemical energy storage.

[0004] In the planning of new power systems, the ability to meet peak-shaving demands in the long term must be included as a constraint and verification factor in the planning. This requires simulating typical operating scenarios with the highest possible peak-shaving demand. However, due to the presence of newly built wind farms and photovoltaic power plants in the planned power grid, the installed capacity of wind and photovoltaic power differs significantly from that of the existing grid, making it impossible to directly generate scenarios from operational data. Existing analytical methods generally require building probabilistic models based on historical wind / solar data, generating numerous wind and solar power output and net load curves through statistical sampling simulations, and then using methods such as clustering to reduce the scenarios, using cluster centers as typical scenarios to analyze peak-shaving demand. This process introduces significant analytical errors in the probabilistic modeling of wind and solar power plants and the sampling simulation stages, significantly reducing the effectiveness of peak-shaving demand assessment.

[0005] This patent proposes a novel method for generating typical scenarios to assess peak-shaving demand. This method can efficiently and more accurately generate typical scenarios that reflect the peak-shaving demand of power systems with a high proportion of renewable energy planning. The proposed method includes two key steps: clustering historical wind / solar / load curves based on the relative peak-valley difference matrix, and generating typical peak-shaving demand scenarios for planning. Applying the patented algorithm can generate typical operating scenarios for assessing peak-shaving demand and constraints in power grid optimization planning, demonstrating clear application needs. Summary of the Invention

[0006] In view of the above-mentioned problems, the present invention is proposed.

[0007] Therefore, the technical problem solved by this invention is that traditional methods rely on probabilistic modeling and random sampling to generate wind and solar power output curves, which leads to large errors in peak-shaving demand analysis and makes it difficult to accurately reflect the actual peak-shaving pressure of the planned power grid. Furthermore, the wind and solar power installed capacity of the planned power grid is different from that of the historically operating power grid, and traditional methods cannot directly use historical data to generate typical peak-shaving scenarios.

[0008] To address the aforementioned technical problems, this invention provides the following technical solution: a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, comprising,

[0009] Collect and preprocess time-series data of the new energy power system; construct an extreme value sequence based on the preprocessed time-series data of the new energy power system to obtain a relative peak-valley difference matrix; perform clustering using the relative peak-valley difference matrix as a feature index to obtain clustering results; extract representative daily time-series data curves of the new energy power system based on the clustering results; combine the time-series data curves of the new energy power system to obtain the representative daily scenario of net load for high peak-shaving demand in the target year.

[0010] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the method involves: normalizing the preprocessed renewable energy power system time-series data and dividing it into time segments.

[0011] The extreme value sequence is calculated for the normalized time series data of the new energy power system according to time segments.

[0012] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the clustering includes:

[0013] Clustering is performed on daily time series samples of new energy power system time series data to obtain sample clusters, that is, the same sample cluster is called the same scenario.

[0014] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the renewable energy power system time-series data includes historical load data, historical wind power data, and historical photovoltaic data, that is, the renewable energy power system time-series data curves include load curves, wind power curves, and photovoltaic curves.

[0015] The extracted representative daily time-series data curves of the new energy power system include...

[0016] Extract representative daily load curves, wind power representative daily curves, and photovoltaic representative daily curves.

[0017] Generate peak demand curves for wind power, photovoltaic power, and typical representative days of load for the target year.

[0018] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the extraction of representative daily load curves includes,

[0019] From each cluster of load samples, select the day with the largest peak-to-valley difference as the representative day load curve.

[0020] Extracting representative daily curves for wind power and photovoltaic power includes,

[0021] For each representative daily load curve extracted, record the period of maximum peak load as i and the period of minimum valley load as j. Then, for each cluster of wind and solar power samples, take element b from the relative peak-valley difference matrix. ij The smallest sample is taken as the representative daily curve for wind power and photovoltaic power of the corresponding cluster.

[0022] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the method for generating typical representative daily curves of wind power, photovoltaic power, and load for high peak-shaving demand in the target year includes:

[0023] Based on the predicted maximum load and planned wind / solar installed capacity for the planning target year, multiply by the normalized representative daily curves of load, wind, and solar extracted in the previous two steps to generate representative daily curves of wind power, solar power, and typical load for the planning target year.

[0024] As a preferred embodiment of the method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning described in this invention, the scenario for obtaining the net load representative day scenario of high peak-shaving demand in the target year includes:

[0025] The representative wind power, photovoltaic, and typical load daily curves of the planning target year are fully combined to generate a set of high peak demand daily operation scenarios that include wind, solar and load daily curves, and the net load curve for each combination is calculated.

[0026] Let L be the cluster of wind power samples, M be the cluster of photovoltaic samples, and N be the cluster of load samples. Then, a total of (L×M×N) peak-shaving demand scenarios can be generated.

[0027] For each scenario, the net load curve {P} is calculated using the following formula. NL,i};

[0028] P NL,i =P LD,i -P WD,i -P PV,i

[0029] In the formula P LD,i PWD,i P PV,i These represent the power values ​​of the i-th 15-minute period of the load, wind power, and photovoltaic power within the same combination, respectively. A total of L×M×N net load curves are obtained, which represent the net load representative day scenarios of L×M×N types of high peak-shaving demand in the target year, i.e., the typical peak-shaving demand scenarios are planned.

[0030] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method for generating typical peak-shaving demand scenarios for planning a high proportion of new energy power systems.

[0031] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the method for generating typical peak-shaving demand scenarios for planning high-proportion renewable energy power systems.

[0032] The beneficial effects of this invention are as follows: This invention proposes the definition of the relative peak-valley difference matrix and uses the relative peak-valley difference matrix as a feature index of the clustering algorithm to group historical sample data of wind / solar / load. This can more effectively group daily power curves with similar peak and valley occurrence periods and similar intraday maximum peak-valley differences into one class.

[0033] Based on the clustering results, considering the impact of renewable energy output on load and the maximum peak-to-valley difference, representative daily wind / solar / load curves are extracted. Then, based on the planned target year's maximum load and renewable energy installed capacity, typical wind / solar / load generation curve scenarios and net load curves for peak-shaving demand assessment are generated. This fully considers the peak-shaving demand boundary scenarios in the long-term operation of the new power system. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, as provided in one embodiment of the present invention.

[0036] Figure 2 This is a diagram illustrating the clustering of wind power curves in a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, provided as an embodiment of the present invention.

[0037] Figure 3 This is a diagram illustrating the clustering of photovoltaic power curves in a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, provided as an embodiment of the present invention.

[0038] Figure 4 This is a load power curve clustering diagram for a method to generate typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, provided as an embodiment of the present invention.

[0039] Figure 5 This invention provides a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, which includes a typical wind / solar / load curve for the target year's peak-shaving demand.

[0040] Figure 6 This invention provides a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, including 18 representative daily net load curves for high peak-shaving demand. Detailed Implementation

[0041] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0042] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, including:

[0043] S1: Collect time-series data of the new energy power system and perform preprocessing.

[0044] In this application embodiment, the time-series data of the new energy power system includes wind power, photovoltaic, and historical load data.

[0045] S2: Construct an extreme value sequence based on the preprocessed time series data of the new energy power system to obtain the relative peak-valley difference matrix.

[0046] In the embodiments of this application, the extreme value sequence is a peak value sequence and a valley value sequence.

[0047] After preprocessing the historical power data of wind, solar and load in the example, the wind power / solar power data is normalized to the annual maximum power generation, and the load data is normalized to the annual maximum load.

[0048] Secondly, the daytime is divided into 6 time intervals, specifically numbered as follows:

[0049] ①: 23:00–3:00; ②: 3:00–7:00;

[0050] ③: 7:00–11:00; ④: 11:00–15:00;

[0051] ⑤: 15:00–19:00; ⑥: 19:00–23:00;

[0052] For the normalized daily wind / solar / load power time series samples, the power peak and trough values ​​for the above 6 time periods are calculated respectively, and the peak sequence X(d) and trough sequence Y(d) for each day are obtained as follows:

[0053] X(d)={P max (i)|i∈[1,6],i∈N}

[0054] Y(d)={P min (i)|i∈[1,6],i∈N}

[0055] Based on the power peak-valley value sequence for each day, a relative peak-valley difference matrix B(d) is constructed as follows:

[0056]

[0057] b i,j =X(d) i -Y(d) j

[0058] =P max (i)-P min (j)

[0059] (i,j∈[1,6] and i,j∈N).

[0060] Among them, P max (i) represents the peak power in the i-th time interval, P min (j) represents the power valley value in the j-th interval.

[0061] S3: Clustering is performed using the relative peak-valley difference matrix as a feature index to obtain the clustering results.

[0062] In the embodiments of this application, the relative peak-to-valley difference matrix B(d) is used as a feature index to cluster the daily time series samples of wind / solar / load power respectively. K-means method or other clustering algorithms can be used to obtain similar sample clusters. The same sample cluster is called the same scenario.

[0063] Since the relative peak-valley difference matrix retains key information such as the peak value, valley value and time distribution of the daily power curve, it can be used as a feature index to more effectively cluster samples with similar peak-shaving needs into the same sample cluster, creating conditions for generating typical peak-shaving scenarios in the subsequent planning stage.

[0064] S4: Extract representative daily time-series data curves of the new energy power system based on the clustering results.

[0065] In this embodiment, based on the clustering results of historical wind / solar / load samples by S3, a calculation method suitable for generating typical peak-shaving demand scenarios in planned power grids is proposed. The main steps are as follows:

[0066] S41: Extract the representative daily load curve.

[0067] From each cluster of load samples, select the day with the largest peak-to-valley difference as the representative day load curve.

[0068] S42: Extract the daily curve representing wind / light.

[0069] For each representative daily load curve extracted in the previous step, let the period of maximum peak load be i and the period of minimum valley load be j. Then, for each cluster of samples from wind power and photovoltaics, take element b from the relative peak-valley difference matrix. ij The smallest sample is used as the representative day curve for wind / light for this cluster.

[0070] S43: Generate typical daily curves for peak demand in the target year, including wind, solar, and load.

[0071] Based on the predicted maximum load and planned wind / solar installed capacity for the planning target year, multiply by the normalized representative daily curves of load, wind, and solar extracted in the previous two steps to generate representative typical daily curves of wind / solar / load for the planning target year.

[0072] S5: Combine the time-series data curves of the new energy power system to obtain the net load representative day scenario of the peak demand in the target year.

[0073] The typical daily curves of wind / solar / load for the planning target year obtained in step S4 are fully combined to generate a set of high peak demand daily operation scenarios that include the daily curves of wind, solar and load, and the net load curve for each combination is calculated.

[0074] Let L be the clusters of wind power samples, M be the clusters of photovoltaic samples, and N be the clusters of load samples. Then, a total of (L×M×N) peak-shaving demand scenarios can be generated. Under each scenario, its net load curve {P} is calculated using the following formula. NL,i}

[0075] P NL,i =P LD,i -PWD,i -P PV,i

[0076] In the formula P LD,i P WD,i P PV,i These represent the power values ​​for the i-th 15-minute period of the same combination of load, wind power, and photovoltaic power within a day. A total of (L×M×N) net load curves are obtained, which represent the typical daily scenarios of net load with (L×M×N) types of high peak-shaving demand in the target year, i.e., the typical peak-shaving demand scenarios planned.

[0077] Example 2, refer to Figures 2-6 This invention provides a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning. To verify the beneficial effects of this invention, scientific demonstration is conducted through experiments.

[0078] The sample data for the example comes from the wind / solar / load power of a province in my country throughout 2022, with a sampling period of 15 minutes. After normalization according to the maximum wind / solar power generation and maximum load, a time series sample of 96 points per day for 365 days is obtained.

[0079] Taking wind power as an example, a relative peak-to-valley difference matrix is ​​constructed for the daily wind power curve samples. Taking day 1 as an example, its peak sequence is as follows:

[0080] X(1)={0.63972,0.59497,0.37942,0.43400,0.45635,0.49569}

[0081] The valley value sequence is as follows:

[0082] Y(1)={0.60800,0.38920,0.22607,0.23378,0.36728,0.38386}

[0083] The relative peak-valley matrix is ​​as follows:

[0084]

[0085] Using the elbow method, the number of clusters was determined to be 3. K-means clustering was performed on the daily power curves of wind power using the relative peak-to-valley difference matrix as a feature index. The results are as follows. Figure 2 As shown.

[0086] Similarly, the photovoltaic curves were clustered into three clusters, such as Figure 3 As shown.

[0087] The load curves were clustered into two clusters, such as Figure 4 As shown.

[0088] Given the predicted maximum load power S in the target year of the planningL =40000 (MW), wind power installed capacity S W =10000 (MW), photovoltaic installed capacity S P =6500 (MW). The peak load period i = 5 and the valley load period j = 2 correspond to the two load power curve clusters. b is selected from the three wind power clusters and the three photovoltaic clusters respectively. ij The smallest sample size, multiplied by its respective installed capacity, yields the representative daily curves of peak demand for wind / solar / load in the planned target year, as shown below. Figure 5 As shown.

[0089] By fully combining the daily wind, solar, and load curves, 3×3×2=18 possible combinations are obtained, which constitute the typical peak-shaving demand scenario set for the planned power grid. The wind / solar / load and net load power curves are shown below. Figure 6 As shown.

[0090] This embodiment also provides an electronic device applicable to a method for generating typical peak-shaving demand scenarios for a high-proportion renewable energy power system plan, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the method for generating typical peak-shaving demand scenarios for a high-proportion renewable energy power system plan as proposed in the above embodiment.

[0091] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a method for generating typical peak-shaving demand scenarios for high-proportion renewable energy power system planning, as proposed in the above embodiment.

[0092] The storage medium proposed in this embodiment and the method for generating typical peak-shaving demand scenarios for planning a high proportion of new energy power systems proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0093] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0094] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A peak shaving demand typical scenario generation method for high-proportion new energy power system planning, characterized in that: The method comprises the following steps: Collecting and preprocessing time series data of a new energy power system; Based on the preprocessed time series data of the new energy power system, an extreme value sequence is constructed to obtain a relative peak-valley difference matrix; Clustering is performed based on the relative peak-valley difference matrix as a characteristic index to obtain a clustering result; According to the clustering result, a representative day new energy power system time series data curve is extracted; The new energy power system time series data curve is combined to obtain a net load representative day scenario of high peak demand in a target year.

2. The method of claim 1, wherein the method is for a high-proportion new energy power system planning. Based on the preprocessed time series data of the new energy power system, normalization is performed, and time segments are divided; The extreme value sequence is calculated according to the time segments of the normalized new energy power system time series data.

3. The method of claim 2, wherein the method is for a high-proportion new energy power system planning. The clustering comprises the following steps: Clustering is performed on the daily time series samples of the new energy power system time series data to obtain sample clusters, i.e., the same sample cluster is called the same scenario.

4. The method of claim 3, wherein the method is used for high-proportion new energy power system planning. The new energy power system time series data comprises historical load data, historical wind power data, and historical photovoltaic data, i.e., the new energy power system time series data curve comprises a load curve, a wind power curve, and a photovoltaic curve; The extraction of the representative day new energy power system time series data curve comprises the following steps: Extraction of a representative day load curve, extraction of a wind power representative day curve, and extraction of a photovoltaic representative day curve; Generation of a wind power, photovoltaic, and load typical representative day curve of high peak demand in a target year.

5. The method of claim 4, wherein the method is used for high-proportion new energy power system planning. The extraction of the representative day load curve comprises the following steps: From each cluster in the load sample cluster, the day with the maximum peak-valley difference is selected as the representative day load curve; The extraction of the wind power representative day curve and the photovoltaic representative day curve comprises the following steps: For each representative daily load curve extracted, record the time period i in which the maximum peak load occurs and the time period j in which the minimum valley load occurs, then for each cluster of samples of wind power and photovoltaic power, take the element b ij in the relative peak-valley difference matrix, take the minimum sample as the representative daily curve of wind power and photovoltaic power of the corresponding cluster.

6. The method of claim 5, wherein the method is for a high-proportion new energy power system planning. The generation of the wind power, photovoltaic, and load typical representative day curve of high peak demand in a target year comprises the following steps: According to the predicted maximum load of the planning target year and the planning wind / photovoltaic installed capacity, the normalized load, wind, and photovoltaic representative day curves extracted in the previous two steps are multiplied to generate the representative wind power, photovoltaic, and load typical representative day curve of the planning target year.

7. The method of claim 6, wherein the method is for a high-proportion new energy power system planning. The generation of the wind power, photovoltaic, and load typical representative day curve of high peak demand in a target year comprises the following steps: The representative wind power, photovoltaic, and load typical representative day curve of the planning target year is completely combined to generate a high peak demand daily operation scenario set containing wind, light, and load day curves, and the net load curve of each combination is calculated; The clustering cluster of the wind power sample is L, the clustering cluster of the photovoltaic sample is M, and the clustering cluster of the load sample is N, so that L×M×N peak shaving demand scenarios can be generated by combination; The net load curve {P NL,i} of each scenario is calculated according to the following formula: P NL,i = P LD,i - P WD,i - P PV,i In the formula, P LD,i , P WD,i , P PV,i respectively represent the power value of the i-th 15 minutes of the load, wind power and photovoltaic in the same combination within one day, and a total of LxMxN groups of net load curves are obtained, that is, LxMxN kinds of high peak demand net load representative day scenarios are generated for the target year, that is, the planning typical peak demand scenario.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to realize the steps of the peak shaving demand typical scenario generation method for high-proportion new energy power system planning in any one of claims 1 to 7.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the peak shaving demand typical scenario generation method for high-proportion new energy power system planning in any one of claims 1 to 7.

Images