Power grid new energy standby demand evaluation method and system considering time sequence characteristics
Through dynamic time regularization and probability density function calculation, the problem that the timing characteristics in the grid's new energy backup demand assessment is not considered, and a more accurate new energy backup demand assessment is achieved, which improves the grid's response capabilities and operational economy.
Patent Information
- Application Number
- CN202410912949.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-09
- Publication Date
- 2024-11-08
AI Technical Summary
The existing technology fails to fully consider the timing characteristics of new energy generation when evaluating the backup demand for new energy in the power grid, resulting in the inaccurate reserve capacity reserves being insufficient and it is difficult to effectively deal with the uncertainty of new energy generation.
Through dynamic time regularization method, screen historical sample data with similar morphology, calculate the probability density function of new energy prediction error, and calculate the backup demand for new energy based on the confidence level to form a systematic evaluation method and system.
It improves the calculation accuracy of new energy backup demand, improves the power grid's ability to respond to uncertainty in new energy generation, and ensures the balance and safe operation of the power grid.
Smart Images

Figure CN118917574A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a method and system for evaluating the demand for new energy standby power grid, and in particular to a method and system for evaluating the demand for new energy standby power grid taking into account time sequence characteristics. Background Art
[0002] With the continuous advancement of the construction of new power systems, the proportion of renewable energy has gradually increased. Its random volatility has brought great challenges to the operation of the power system. In order to ensure the real-time balance of power sources and loads in the power grid, reserving backup for renewable energy is a necessary measure to cope with the strong uncertainty of its power generation.
[0003] Traditional power systems need to maintain a certain reserve capacity to ensure safe and stable operation of the system and cope with unexpected load fluctuations, failures, etc. The total reserve capacity of the power system is usually determined by a deterministic method, such as a percentage of the system's annual maximum load or total installed capacity, or the capacity of the system's largest unit. This method usually takes the safe and stable operation of the system as the primary consideration, and rarely considers the impact of new energy uncertainties on the total reserve capacity.
[0004] In order to cope with the uncertainty of renewable energy generation, the power grid control center usually uses the predicted power multiplied by the discount coefficient to reserve renewable energy reserve capacity. However, it is difficult to accurately describe the rationality of the reserved reserve by simply assigning the discount coefficient based on manual experience. With the large-scale grid connection of renewable energy, the method of determining the reserve capacity by probabilistic statistics of the historical forecast error data of renewable energy has also been studied in the literature, but it lacks consideration of the time series characteristics. Therefore, how to consider the time series characteristics of the historical data of renewable energy forecasts and conduct the assessment of the renewable energy reserve demand of the power grid has become a technical problem that technicians in this field urgently need to solve. Summary of the invention
[0005] Purpose of the invention: The purpose of the present invention is to provide a method and system for evaluating the new energy standby demand of power grids taking into account time series characteristics. By measuring the morphological similarity of time series to screen historical sample data and statistically analyzing the error distribution of historical predictions of new energy, a method for evaluating the new energy standby demand of power grids is proposed, which helps to improve the ability of the power grid to cope with the uncertainty of renewable energy power generation.
[0006] Technical solution: The method for evaluating the demand for new energy backup power grid considering the timing characteristics of the present invention comprises the following steps:
[0007] A new energy forecast curve of a historical date similar in shape to the new energy forecast curve of the planned date is obtained by a dynamic time warping method, and the historical date is added to the historical sample date set;
[0008] According to the date in the historical sample date set, the difference between the new energy prediction curve and the new energy measured curve in each period of the date is calculated to generate a historical sample data sequence of new energy prediction error;
[0009] Cluster the time periods in the historical sample data sequence to obtain aggregated time periods;
[0010] Calculate the probability density function of the new energy forecast error for the aggregation period;
[0011] The new energy reserve demand is calculated according to the set confidence level and the probability density function.
[0012] Furthermore, the new energy forecast curve of a historical date similar in form to the new energy forecast curve of the planned date is obtained by the dynamic time warping method, and the historical date is used as a historical sample date set including:
[0013] Construct an n×n dimensional distance matrix D, whose element d ij for:
[0014] d ij =(x i -y j ) 2 i,j∈[1,n];
[0015] Among them, x i is the new energy forecast value for the planned date period i, y j is the predicted value of new energy in a certain day period j in history, d ij is the Euclidean distance between the new energy forecast value of the planned date period i and the new energy forecast value of a historical date period j;
[0016] Construct a path vector W=[w1,w2,...,w l ], whose element w p for:
[0017] w p =d ij p∈[1,l];
[0018] Among them, w p is the pth path of the dynamic time warping path;
[0019] Dynamic Time Warping Distance D w The calculation formula is:
[0020]
[0021] Where the time series X = [x1, x2, ..., x n ] is the new energy forecast curve for the planned date, x nis the new energy forecast value for the planned date period n, time series Y = [y1, y2, ..., y n ] is the new energy prediction curve for a certain day in history, y n is the new energy forecast value for the historical date period n;
[0022] If the dynamic time warping distance between the new energy forecast curve of the historical date and the new energy forecast curve of the planned date is less than a threshold, the historical date is used as the historical sample date.
[0023] Furthermore, the step of calculating the difference between the new energy prediction curve and the new energy measured curve in each period of the date according to the date in the historical sample date set to generate a historical sample data sequence of new energy prediction error includes:
[0024] The new energy prediction error is the new energy prediction value minus the new energy measured value:
[0025] P err,t =P fort,t -P real,t ;
[0026] Among them, P err,t is the new energy prediction error in period t, P fort,t is the predicted value of new energy in period t, P real,t is the measured value of new energy in period t;
[0027] The historical sample data sequence of new energy forecast error on any historical sample date is:
[0028] Z=[P err,1 ,P err ,2,...,P err,N ];
[0029] Among them, Z is the new energy forecast error sample data sequence of the historical sample date, N is the number of time periods of the historical sample date, P err,N is the new energy prediction error in period N.
[0030] Furthermore, clustering the time periods in the historical sample data sequence to obtain the aggregated time periods includes:
[0031] (1) Perform hierarchical clustering on the output power sequence of each period in the historical sample data sequence to obtain K partitions, calculate the centers of the K partitions, and use them as the initial clustering centers of the clustering algorithm;
[0032] (2) For each time period, the Euclidean distance between the output power sequence and the corresponding K cluster centers is calculated, and the sequence is assigned to the cluster with the shortest distance to form new K clusters.
[0033] (3) For the newly formed K clusters, the cluster center of each cluster is recalculated using the mean method;
[0034] (4) Repeat steps (2) and (3) until the class members no longer change. The iteration ends and the aggregation period is output.
[0035] Further, the probability density function of the new energy prediction error in the calculation aggregation period includes: the probability density function obeys a normal distribution.
[0036] Furthermore, the probability density function of the new energy prediction error during the calculation aggregation period includes:
[0037] The probability density function is:
[0038]
[0039]
[0040]
[0041] Among them, μ k is the expected distribution of the new energy forecast error for time period k, M is the number of historical sample dates, K is the number of time period ranges divided into one day after time period aggregation, T k is the number of time periods included in the time period range k, P err,i is the prediction error data of the i-th new energy, σ k is the variance of the new energy forecast error distribution in time period k; x is the variable of the new energy forecast error, f k (x) is the probability density function of the new energy prediction error in time period k.
[0042] Further, the calculating the new energy reserve demand according to the set confidence level and the probability density function includes:
[0043] The calculation method of the new energy reserve demand is:
[0044]
[0045] Among them, R w,k is the new energy reserve demand in time range k, α is the confidence level of the new energy reserve demand, and f k (x) is the probability density function of the new energy prediction error in time period k.
[0046] The grid new energy standby demand assessment system considering timing characteristics of the present invention comprises:
[0047] A historical sample date set generation module is used to obtain a new energy forecast curve of a historical date that is similar in shape to the new energy forecast curve of the planned date through a dynamic time warping method, and add the historical date to the historical sample date set;
[0048] A historical sample data sequence generation module is used to calculate the difference between the new energy prediction curve and the new energy measured curve in each period of the date according to the date in the historical sample date set, and generate a historical sample data sequence of new energy prediction error;
[0049] The module for aggregating time periods with similar forecast errors is used to cluster the time periods in the historical sample data sequence to obtain aggregated time periods;
[0050] A probability density function calculation module is used to calculate the probability density function of the new energy prediction error during the aggregation period;
[0051] The new energy standby demand calculation module is used to calculate the new energy standby demand according to the set confidence level and the probability density function.
[0052] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and is characterized in that when the computer program is loaded into the processor, the method for evaluating the new energy standby demand of a power grid considering timing characteristics is implemented.
[0053] The computer-readable storage medium described in the present invention stores a computer program, and when the computer program is executed by a processor, it implements the method for evaluating the new energy standby demand of the power grid considering the timing characteristics.
[0054] Beneficial effects: Compared with the prior art, the advantages of the present invention are: the method and system for evaluating the backup demand of new energy in the power grid taking into account the time series characteristics provided by the present invention are helpful to make full use of the time series variation characteristics of the historical data of new energy prediction, realize the accurate characterization of the probability distribution of the new energy prediction error, improve the calculation accuracy of the backup demand of new energy, and improve the balancing ability of the power grid and the level of new energy consumption. Specifically:
[0055] 1) The dynamic time warping method is used to calculate the morphological similarity by matching and mapping the time series, which can effectively advance the historical data similar to the new energy forecast time series data of the planned day.
[0056] 2) Considering the forecast error and volatility of renewable energy, calculating the grid's renewable energy backup demand can enhance the grid's ability to cope with the uncertainty of renewable energy generation.
[0057] 3) Conduct statistical analysis on historical data of new energy prediction errors to provide more accurate new energy backup demand and improve the economic efficiency of system operation while ensuring the safety of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS
[0058] Figure 1 This is a flow chart of the new energy reserve demand assessment method of the present invention;
[0059] Figure 2 This is a diagram of the architecture of the new energy standby demand assessment system of the present invention. DETAILED DESCRIPTION
[0060] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.
[0061] like Figure 1 As shown, the method for evaluating the demand for new energy backup power grid considering time sequence characteristics comprises the following steps:
[0062] Step 1: Use the dynamic time warping method to measure the new energy forecast curve of the historical date that is similar to the time series shape of the new energy forecast curve of the planned date as the historical sample date set.
[0063] (1.1) Let the time series X = [x1, x2, ..., x n ] is the new energy forecast curve for the planned date, x n is the new energy forecast value for the planned date period n, time series Y = [y1, y2, ..., y n ] is the new energy prediction curve for a certain day in history, y n is the new energy forecast value for the historical date period n, then an n×n dimensional distance matrix D is constructed, whose element d ij for:
[0064] d ij =(x i -y j ) 2 i,j∈[1,n];
[0065] Among them, x i is the new energy forecast value for the planned date period i, y j is the predicted value of new energy in a certain day period j in history, d ij It is the Euclidean distance between the new energy forecast value of the planned date period i and the new energy forecast value of a historical date period j.
[0066] (1.2) Construct the path vector W = [w1, w2, ..., w l ], whose element w p for:
[0067] w p =d ij p∈[1,l];
[0068] Among them, wp is the pth path of the dynamic time warping path.
[0069] (1.3) Dynamic time warping distance D between time series X and Y w The calculation formula is:
[0070]
[0071] (1.4) A threshold value of the dynamic time warping distance is given. If the dynamic time warping distance between the new energy forecast curve of the historical date and the planned date is less than the threshold value, they are similar and can be used as historical sample dates.
[0072] Step 2: Based on the historical sample date set, obtain the new energy forecast curve and the new energy measured curve for each time period, and generate a historical sample data sequence of new energy forecast errors.
[0073] The new energy prediction error is the new energy prediction value minus the new energy measured value:
[0074] P err,t =P fort,t -P real,t ;
[0075] Among them, Perr,t is the prediction error of new energy in period t, Pfort,t is the predicted value of new energy in period t, and Preal,t is the measured value of new energy in period t;
[0076] The historical sample data sequence of new energy forecast error on a certain historical sample date is:
[0077] Z=[P err,1 ,P err,2 ,...,P err,N ];
[0078] Among them, Z is the new energy prediction error sample data sequence of the historical sample date, N is the number of time periods of the historical sample date, and Perr,N is the new energy prediction error of time period N.
[0079] Step 3: Cluster similar time periods of the historical sample data series to obtain time period aggregation results.
[0080] (3.1) Perform hierarchical clustering on the output power sequence of each period in the historical sample data to obtain K partitions, calculate the centers of the K partitions, and use them as the initial cluster centers of the clustering algorithm;
[0081] (3.2) For each time period, calculate the Euclidean distance of the K cluster centers to which it corresponds, and assign it to the cluster with the shortest distance to form new K clusters.
[0082] (3.3) For the formation of new K clusters, the cluster center of each cluster is recalculated using the mean method;
[0083] (3.4) Repeat steps (3.2) and (3.3) until the class members no longer change. The iteration ends and similar clustering time periods are output.
[0084] Step 4: Calculate the probability distribution of new energy forecast errors in each time period after statistical period aggregation to obtain the probability density function.
[0085] The historical sample date set has a total of M days. After period aggregation, one day is divided into K period ranges, and the number of periods contained in each time range is T. k . Then for the kth time period, there are M*T k The error data of new energy prediction output are assumed to obey the normal distribution. The expectation and variance of the error distribution can be obtained by using the maximum likelihood estimation method:
[0086] The expected estimated value of the error in the predicted output of new energy is:
[0087]
[0088] The estimated variance of the new energy output forecast error is:
[0089]
[0090] Among them, μ k is the expected distribution of the error of the new energy output forecast in the time range k, M is the number of historical sample dates, K is the number of time ranges divided into one day after the time range aggregation, T k is the number of time periods included in the time period range k, P err,i is the prediction error data of the i-th new energy, σ k is the variance of the error distribution of the new energy output forecast within time period k.
[0091] The probability density function of the error in the predicted output of new energy is:
[0092]
[0093] Among them, x is the variable of the new energy output prediction error, f k (x) is the probability density function of the predicted output error of new energy in time period k.
[0094] Step 5: According to the set confidence level, based on the probability density function of the new energy prediction error, the corresponding backup demand is obtained.
[0095]
[0096]
[0097] Among them, R w,k is the new energy reserve demand in time range k, and α is the confidence level of the new energy reserve demand. The new energy reserve demand can be used as a constraint in the power generation plan on the planned date.
[0098] The grid new energy standby demand assessment system considering timing characteristics of the present invention comprises:
[0099] A historical sample date set generation module is used to obtain a new energy forecast curve of a historical date that is similar in shape to the new energy forecast curve of the planned date through a dynamic time warping method, and add the historical date to the historical sample date set;
[0100] A historical sample data sequence generation module is used to calculate the difference between the new energy prediction curve and the new energy measured curve in each period of the date according to the date in the historical sample date set, and generate a historical sample data sequence of new energy prediction error;
[0101] The module for aggregating time periods with similar forecast errors is used to cluster the time periods in the historical sample data sequence to obtain aggregated time periods;
[0102] A probability density function calculation module is used to calculate the probability density function of the new energy prediction error during the aggregation period;
[0103] The new energy standby demand calculation module is used to calculate the new energy standby demand according to the set confidence level and the probability density function.
[0104] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and is characterized in that when the computer program is loaded into the processor, the method for evaluating the new energy standby demand of a power grid considering timing characteristics is implemented.
[0105] The computer-readable storage medium described in the present invention stores a computer program, and when the computer program is executed by a processor, it implements the method for evaluating the new energy standby demand of the power grid considering the timing characteristics.
[0106] The computer-readable storage media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory, or any other medium that can be used to store the desired program code in the form of instructions or data structures and that can be accessed by a computer.
[0107] The processor is used to execute the computer program stored in the memory to implement each step of the method involved in the above embodiment.
Claims
1. A method for evaluating the demand for new energy backup in a power grid considering time sequence characteristics, characterized in that: The steps include: A new energy forecast curve of a historical date similar in shape to the new energy forecast curve of the planned date is obtained by a dynamic time warping method, and the historical date is added to the historical sample date set; According to the date in the historical sample date set, the difference between the new energy prediction curve and the new energy measured curve in each period of the date is calculated to generate a historical sample data sequence of new energy prediction error; Cluster the time periods in the historical sample data sequence to obtain aggregated time periods; Calculate the probability density function of the new energy forecast error for the aggregation period; The new energy reserve demand is calculated according to the set confidence level and the probability density function.
2. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1 is characterized in that: The new energy forecast curve of the historical date similar in shape to the new energy forecast curve of the planned date is obtained by the dynamic time warping method, and the historical date is used as the historical sample date set including: Construct an n×n dimensional distance matrix D, whose element d ij for: d ij =(x i -y j ) 2 i,j∈[1,n]; Among them, x i is the new energy forecast value for the planned date period i, y j is the predicted value of new energy in a certain day period j in history, d ij is the Euclidean distance between the new energy forecast value of the planned date period i and the new energy forecast value of a historical date period j; Construct a path vector W=[w1,w2,...,w l ], whose element w p for: w p =d ij p∈[1,l]; Among them, w p is the pth path of the dynamic time warping path; Dynamic Time Warping Distance D w The calculation formula is: Where the time series X = [x1, x2, ..., x n ] is the new energy forecast curve for the planned date, x n is the new energy forecast value for the planned date period n, time series Y = [y1, y2, ..., y n ] is the new energy prediction curve for a certain day in history, y n is the new energy forecast value for the historical date period n; If the dynamic time warping distance between the new energy forecast curve of the historical date and the new energy forecast curve of the planned date is less than a threshold, the historical date is used as the historical sample date.
3. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1 is characterized in that: The step of calculating the difference between the new energy prediction curve and the new energy measured curve in each period of the date according to the date in the historical sample date set to generate a historical sample data sequence of new energy prediction error includes: The new energy prediction error is the new energy prediction value minus the new energy measured value: P err,t =P fort,t -P real,t ; Among them, P err,t is the new energy prediction error in period t, P fort,t is the predicted value of new energy in period t, P real,t is the measured value of new energy in period t; The historical sample data sequence of new energy forecast error on any historical sample date is: Z=[P err,1 ,P err ,2,...,P err,N ]; Among them, Z is the new energy forecast error sample data sequence of the historical sample date, N is the number of time periods of the historical sample date, P err,N is the new energy prediction error in period N.
4. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1 is characterized in that: The clustering of time periods in the historical sample data sequence to obtain aggregated time periods includes: (1) Perform hierarchical clustering on the output power sequence of each period in the historical sample data sequence to obtain K partitions, calculate the centers of the K partitions, and use them as the initial clustering centers of the clustering algorithm; (2) For each time period, the Euclidean distance between the output power sequence and the corresponding K cluster centers is calculated, and the sequence is assigned to the cluster with the shortest distance to form new K clusters. (3) For the newly formed K clusters, the cluster center of each cluster is recalculated using the mean method; (4) Repeat steps (2) and (3) until the class members no longer change. The iteration ends and the aggregation period is output.
5. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1, characterized in that: The calculation of the probability density function of the new energy prediction error in the aggregation period includes: the probability density function obeys a normal distribution.
6. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1, characterized in that: The probability density function of the new energy prediction error during the calculation aggregation period includes: The probability density function is: Among them, μ k is the expected distribution of the new energy forecast error for time period k, M is the number of historical sample dates, K is the number of time period ranges divided into one day after time period aggregation, T k is the number of time periods included in the time period range k, P err,i is the prediction error data of the i-th new energy, σ k is the variance of the new energy forecast error distribution in time period k; x is the variable of the new energy forecast error, f k (x) is the probability density function of the new energy prediction error in time period k.
7. The method for evaluating the demand for new energy backup power grid considering timing characteristics according to claim 1, characterized in that: The calculating of the new energy reserve demand according to the set confidence level and the probability density function includes: The calculation method of the new energy reserve demand is: Among them, R w,k is the new energy reserve demand in time range k, α is the confidence level of the new energy reserve demand, and f k (x) is the probability density function of the new energy prediction error in time period k.
8. A grid new energy standby demand assessment system considering timing characteristics, characterized in that: include: A historical sample date set generation module is used to obtain a new energy forecast curve of a historical date that is similar in shape to the new energy forecast curve of the planned date through a dynamic time warping method, and add the historical date to the historical sample date set; A historical sample data sequence generation module is used to calculate the difference between the new energy prediction curve and the new energy measured curve in each period of the date according to the date in the historical sample date set, and generate a historical sample data sequence of new energy prediction error; The module for aggregating time periods with similar forecast errors is used to cluster the time periods in the historical sample data sequence to obtain aggregated time periods; A probability density function calculation module is used to calculate the probability density function of the new energy prediction error during the aggregation period; The new energy standby demand calculation module is used to calculate the new energy standby demand according to the set confidence level and the probability density function.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into the processor, it implements the method for evaluating the new energy reserve demand of the power grid considering the timing characteristics according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for evaluating the new energy reserve demand of a power grid considering timing characteristics according to any one of claims 1 to 7 is implemented.
Citation Information
Cited By
Operation reserve capacity demand evaluation method and system considering uncertainty of new energy
CN120494451A
Method and system for evaluating operational reserve capacity requirements considering uncertainty of new energy sources
CN120494451B
