A step time-of-use electricity price optimization method and terminal

Abstract

The application discloses a kind of step time-of-use electricity price optimization method and terminal, first, by clustering algorithm determines electricity price number and the electricity corresponding to each number, can make each grade electricity limit more in line with actual situation, improve its objectivity and applicability;Further, according to Ramsey pricing is priced for the electricity of each number, using moving boundary technology on time-of-use electricity price optimization, can effectively improve the accuracy of time period division;And with the minimum peak load and peak valley difference as target setting target function, in combination with time period and the price setting constraint condition of each number, establish step time-of-use electricity price optimization model, while not affecting the power demand of daily life of resident, can effectively guarantee the smooth implementation of peak clipping and valley filling strategy, to guide resident reasonable electricity, realize the balanced allocation of electric power load.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity price optimization, and in particular to an optimization method and terminal for a stepped time-of-use electricity price. Background Art

[0002] With the deepening of electricity market reforms, the continued growth of residential electricity demand, and changes in load characteristics, traditional residential electricity pricing policies have gradually revealed their limitations. While some progress has been made in stabilizing electricity prices and promoting energy conservation and emission reduction, the increasing openness of the electricity market and the diversification of residential electricity consumption behavior have made the existing electricity pricing mechanism difficult to adapt to new development needs. Therefore, adjusting electricity pricing policies, particularly the introduction of a time-of-use electricity price combined with a tiered pricing mechanism, has become a key approach to addressing these current challenges.

[0003] In the research of residential electricity pricing schemes, tiered pricing and tiered time-of-use pricing are two major research areas. Tiered pricing encourages residents to conserve electricity by setting different price levels, but determining the energy split points has always been a technical challenge. Various technologies have been used to analyze and determine the energy splits for tiered pricing from economic and statistical perspectives, but incomplete considerations in modeling often lead to errors in the fitting results.

[0004] In the design of tiered time-of-use electricity pricing, existing technologies attempt to analyze users' differentiated price elasticity through utility functions to determine the optimal price range for each tier of the time-of-use tiered electricity price. However, these time periods are not integrated with actual load data of residents, resulting in a lack of objective basis for setting time periods and a failure to truly reflect actual local load levels. Furthermore, the determination of peak and off-peak periods often relies on the experience of other regions, lacking scientific and targeted approaches. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method and terminal for optimizing stepped time-of-use electricity prices, which can effectively reflect the time value of power resources while ensuring the objectivity of electricity price optimization and achieve guidance for energy conservation and peak regulation.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for optimizing a stepped time-of-use electricity price comprises the following steps:

[0008] S1. Determine the number of electricity price levels and the corresponding electricity volume for each level through a clustering algorithm;

[0009] S2. Set the price of electricity for each level according to Ramsey pricing;

[0010] S3. Initializing the boundary based on a load sequence containing load values ​​for each time period in a typical day, updating the boundary using a moving boundary method, and dividing the time period based on the relationship between the updated boundary and the load values ​​in the load sequence;

[0011] S4. Constraints are set in combination with the time period and the electricity price of each gear, and an objective function is set with the goal of minimizing peak load and peak-to-valley difference. A stepped time-of-use electricity price optimization model is established in combination with the constraints and the objective function, and a stepped time-of-use electricity price scheme is obtained by solving the stepped time-of-use electricity price optimization model.

[0012] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0013] A terminal for optimizing stepped time-of-use electricity prices comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the above-mentioned method for optimizing stepped time-of-use electricity prices is implemented.

[0014] The beneficial effects of the present invention are: first, by determining the number of electricity price levels and the amount of electricity corresponding to each level through a clustering algorithm, the limit of the amount of electricity in each level can be made more in line with the actual situation, thereby improving its objectivity and applicability; further, pricing is performed for the electricity price of each level according to Ramsey pricing, and the moving boundary technology is used in the time-of-use electricity price optimization, which can effectively improve the accuracy of time period division; and the objective function is set with the goal of minimizing peak load and peak-valley difference, and the constraint conditions are set in combination with the time period and the electricity price of each level to establish a stepped time-of-use electricity price optimization model. While not affecting the electricity demand of residents in their daily lives, it can effectively ensure the smooth implementation of the peak shaving and valley filling strategy, thereby guiding residents to use electricity rationally and achieving a balanced distribution of power load. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flow chart of a method for optimizing a stepped time-of-use electricity price according to an embodiment of the present invention;

[0016] Figure 2 A schematic diagram of an optimization terminal for a tiered time-of-use electricity price according to an embodiment of the present invention;

[0017] Figure 3 This is a clustering flow chart of an embodiment of the present invention;

[0018] Figure 4 This is a graph showing the sum of squared errors under different K values ​​in an embodiment of the present invention;

[0019] Figure 5 This is a diagram showing the time period division results of a calculation example according to an embodiment of the present invention;

[0020] Description of labels:

[0021] 1. An optimization terminal for tiered time-of-use electricity prices; 2. Memory; 3. Processor. DETAILED DESCRIPTION

[0022] To illustrate the technical content, achieved objectives and effects of the present invention in detail, the following description is given in conjunction with the embodiments and accompanying drawings.

[0023] Please refer to Figure 1 , an embodiment of the present invention provides a method for optimizing a ladder time-of-use electricity price, comprising the steps of:

[0024] S1. Determine the number of electricity price levels and the corresponding electricity volume for each level through a clustering algorithm;

[0025] S2. Set the price of electricity for each level according to Ramsey pricing;

[0026] S3. Initializing the boundary based on a load sequence containing load values ​​for each time period in a typical day, updating the boundary using a moving boundary method, and dividing the time period based on the relationship between the updated boundary and the load values ​​in the load sequence;

[0027] S4. Constraints are set in combination with the time period and the electricity price of each gear, and an objective function is set with the goal of minimizing peak load and peak-to-valley difference. A stepped time-of-use electricity price optimization model is established in combination with the constraints and the objective function, and a stepped time-of-use electricity price scheme is obtained by solving the stepped time-of-use electricity price optimization model.

[0028] From the above description, it can be seen that the beneficial effects of the present invention are: first, the number of electricity price levels and the amount of electricity corresponding to each level are determined by a clustering algorithm, which can make the boundaries of the electricity amount of each level more in line with the actual situation and improve its objectivity and applicability; further, the electricity price of each level is priced according to Ramsey pricing, and the moving boundary technology is used in the time-of-use electricity price optimization, which can effectively improve the accuracy of time period division; and the objective function is set with the goal of minimizing peak load and peak-valley difference, and the constraints are set in combination with the time period and the electricity price of each level to establish a stepped time-of-use electricity price optimization model. While not affecting the residents' daily electricity demand, it can effectively ensure the smooth implementation of the peak shaving and valley filling strategy, thereby guiding residents to use electricity rationally and achieve balanced distribution of power load.

[0029] Furthermore, step S1 includes:

[0030] S11. Obtain a sample set of power data, calculate the number of clusters using the elbow method, randomly select a sample from the sample set as the first cluster center, and use the remaining samples farthest from the first cluster center as new second cluster centers, until the sum of the number of the first cluster centers and the second cluster centers reaches the specified number of clusters;

[0031] S12, calculating the distances of the remaining samples to the first cluster center or the second cluster center in sequence, and classifying the remaining samples into the cluster corresponding to the cluster center with the shortest distance;

[0032] S13, recalculating the center point of each cluster;

[0033] S14. Repeat steps S12 and S13 until the center points of the clusters no longer change, then output the clustering results, determine the number of electricity price levels based on the number of clusters in the clustering results, and determine the electricity quantity boundary value corresponding to each level based on the cluster boundaries of the clustering results.

[0034] From the above description, we can see that the cluster centers of the cluster number are selected one by one, and the farther the sample point is from the center of other clusters, the greater the possibility of being selected as the cluster center. Finally, the center points of each cluster are reselected. In this way, the rationality of the selection of each cluster and its cluster center is improved.

[0035] Furthermore, step S2 includes:

[0036] Set a preset electricity price for the first tier of electricity prices;

[0037] Pricing for electricity prices in tier 2 and above based on Ramsey pricing:

[0038]

[0039] Where p i represents the average electricity price of the i-th tier, M c represents the long-term marginal cost, η i represents the intensity of demand response in the i-th tier, and R represents the Ramsey index.

[0040] From the above description, it can be seen that the use of Ramsey pricing theory in the process of electricity price pricing not only effectively protects the social welfare of residential users, but also minimizes the negative impact on the allocation of electricity resources, thereby achieving fair, efficient and sustainable development of the electricity market.

[0041] Furthermore, step S3 includes:

[0042] S31. Input a first load sequence of a typical day, sort the first load sequence from low to high, and obtain a second load sequence, where the load sequence includes hourly load values;

[0043] S32: Initialize a first boundary and a second boundary according to the second load sequence, wherein the first boundary is smaller than the second boundary;

[0044] S33, iteratively updating the first boundary and the second boundary by moving the boundary until the iteration result meets a preset condition;

[0045] S34. Compare the load value of each hour in the second load sequence with the first boundary and the second boundary of the updated iteration. If the load value is less than or equal to the first boundary, the hour corresponding to the load value is the valley period. If the load value is greater than the first boundary and less than the second boundary, the hour corresponding to the load value is the normal period. If the load value is greater than the second boundary, the hour corresponding to the load value is the peak period.

[0046] From the above description, it can be seen that by changing the positions of the first boundary and the second boundary to solve the optimal peak-valley time period division method, the intra-class and inter-class distances between different peak-valley time periods are fully considered, thereby improving the accuracy of time period division.

[0047] Furthermore, step S4 includes:

[0048] The objective function is established with the goal of minimizing peak load and minimizing peak-to-valley difference:

[0049]

[0050]

[0051]

[0052] In the formula, M1 represents minimizing the daily load peak, M2 represents minimizing the daily load peak-to-valley difference, and L i,max , L i,min , L' i,max , L' i,min They represent the maximum and minimum daily load values ​​of the i-th electricity consumption segment before electricity price adjustment and the maximum and minimum daily load values ​​after adjustment respectively;

[0053] The constraints of electricity price, time period and quantity are set, and a step-by-step time-of-use electricity price optimization model is established in combination with the objective function and the constraints.

[0054] As can be seen from the above description, reducing peak load and increasing valley load will lead to a smaller peak-valley difference. In other words, any combination of the two can ensure the realization of the other. Therefore, minimizing peak load and minimizing peak-valley difference are selected as the objective function, while avoiding deviation caused by a large difference between the two values.

[0055] Please refer to Figure 2 Another embodiment of the present invention provides an optimization terminal for a stepped time-of-use electricity price, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the processor implements the various steps of the above-mentioned method for optimizing a stepped time-of-use electricity price.

[0056] The above-mentioned method and terminal for optimizing the tiered time-of-use electricity price of the present invention are suitable for effectively reflecting the time value of power resources and guiding energy conservation and peak load regulation while ensuring the objectivity of electricity price optimization. The following is an explanation of the specific implementation methods:

[0057] Example 1

[0058] Please refer to Figure 1 , a method for optimizing a ladder time-of-use electricity price, comprising the steps of:

[0059] S1. Determine the number of electricity price levels and the amount of electricity corresponding to each level through a clustering algorithm.

[0060] The clustering algorithm used in this embodiment is the K-means++ clustering algorithm. The k-means algorithm, as a typical unsupervised machine learning data classification method, can randomly select K sample points as cluster centers when faced with a data set with unknown data distribution, calculate the Euclidean distances from the remaining sample points to each cluster center, and when the sum of the squared distances of the samples within each cluster reaches a minimum, divide the data set into k different clusters, so that the similarity of samples within the same cluster is high, and the similarity of samples between different clusters is low. However, the initial cluster center selection of this algorithm is random, and the convergence situation is heavily dependent on the selection of cluster centers. When the initial location of the cluster center is not selected properly, it is easy to form a local optimal solution instead of a global optimal solution.

[0061] Based on this, this embodiment adopts the k-means++ algorithm, which optimizes the selection method of cluster centers. By selecting k cluster centers one by one, the sample points that are farther away from other cluster centers have a greater probability of being selected as cluster centers. Please refer to Figure 3 , the specific steps are as follows:

[0062] S11. Obtain a sample set of power data, calculate the number of clusters using the elbow method, randomly select a sample from the sample set as the first cluster center, and use the remaining samples farthest from the first cluster center as the new second cluster center, until the sum of the number of the first cluster centers and the second cluster centers reaches the number of clusters.

[0063] Based on the actual monthly electricity consumption data of residents, the number of clusters can be calculated by using the k-means++ algorithm and calculating the sum of the squared errors (SSE) through the elbow method. The optimal k value in this embodiment is 3. Please refer to Figure 4 , that is, adopting a three-tier electricity price model.

[0064] Specifically, randomly select a sample from the n sample sets as the cluster center c1, and calculate the remaining sample points x in turn. i The distance d(x i,c1), then the probability of the remaining sample points being selected as the next cluster center is:

[0065]

[0066] The next cluster center is selected by the roulette wheel algorithm. Repeat this step until k cluster centers are selected and proceed to step S12.

[0067] S12. Calculate the distances of the remaining samples to the center of the first cluster or the center of the second cluster in sequence, and classify the remaining samples into the cluster corresponding to the cluster center with the shortest distance. Specifically, calculate the distances of the remaining samples to each cluster center in sequence, and classify the samples into the cluster with the shortest distance.

[0068] S13. Recalculate the center point of each cluster.

[0069] S14. Repeat steps S12 and S13 until the center points of the clusters no longer change, then output the clustering results, determine the number of electricity price levels based on the number of clusters in the clustering results, and determine the electricity quantity boundary value corresponding to each level based on the cluster boundaries of the clustering results.

[0070] By clustering historical electricity consumption data, residential users with similar electricity consumption are grouped together, ultimately forming multiple user clusters. Similar electricity consumption levels within a specific period often reveal similar electricity usage behavior, which in turn reflects various potential influencing factors. The multiple clusters derived from cluster analysis serve as a natural basis for tiered electricity consumption classification. Specifically, the number of clusters directly determines the number of segments in the tiered electricity consumption, while the boundaries between adjacent clusters serve as reference values ​​for the electricity limits separating the tiered segments. This classification approach is both data-driven and objective, while also reflecting the actual differences in user electricity consumption behavior.

[0071] S2. Set the price of electricity for each level according to Ramsey pricing.

[0072] In theory, conventional commodity prices should offset costs, enabling businesses to operate normally, while also generating a certain profit. However, due to the nature of residential electricity prices as public utility products, current residential electricity prices are far below the cost of supplying electricity. To ensure that residential electricity prices maximize energy efficiency while meeting cost-compensation constraints, this embodiment determines the average residential electricity price levels for each tier based on Ramsey pricing theory. Ramsey pricing sets different prices for different income groups. The basic principle is to set higher prices for users with lower price elasticity of demand and lower prices for users with lower price elasticity. The goal is to ensure that businesses balance their revenues and expenditures and maximize social welfare.

[0073] Since household income is positively correlated with the number of electricity consumption segments, higher income households consume more electricity and are less price-sensitive. Therefore, this theory is applied to residential electricity pricing by implementing a stepped-up pricing system for each market segment, from low to high electricity consumption. The rules for setting each price tier are as follows: the first tier is for basic electricity consumption, aiming to meet the basic living electricity needs of low-income groups; the second tier is for normal and reasonable electricity consumption, aiming to meet the normal living electricity needs of all residents; and the third tier and above are for luxury electricity consumption by higher-income groups. The price increase for each tier should adhere to certain constraints: the first tier should remain consistent with the current price or increase slightly; the second tier should gradually transition to the power supply cost based on residents' affordability to reduce leakage effects; and the third tier and above should see significant price increases to reduce unintentional energy waste by residents, improve energy efficiency, and further alleviate the pressure of cross-subsidy from industry and commerce to residents.

[0074] In this embodiment, a preset electricity price is set for the first tier of electricity prices; the second tier and above electricity prices should achieve optimal allocation of electric energy resources while satisfying the cost compensation constraint. That is, the second tier and above electricity prices are priced according to Ramsey pricing, and the relationship formula between the data in different tiers is provided:

[0075]

[0076]

[0077] Where M c represents the long-term marginal cost; η i ,η j Indicates the intensity of demand response at the i-th and j-th levels, which is generally a constant; p i 、p j represents the electricity price for the i-th tier user, q i represents the total electricity consumption of the i-th gear, and C represents the power supply cost.

[0078] The power demand function of each gear can be expressed as:

[0079] ln(q i )=η i ln(p i )+ln(k i )

[0080] Where k i It is the exponent of the segmented demand function of electricity consumption in the i-th tier, which is generally a constant.

[0081] Introducing the Ramsey index R = λ / (1+λ), λ is the Lagrangian factor, and the demand response of each gear is η i As a constant, the formula is rearranged to obtain the average electricity price level of the i-th tier:

[0082]

[0083] In the formula, i = 2, 3,…, n, where n is the total number of tiers in the tiered electricity price.

[0084] According to Ramsey's second-best pricing, which aims to ensure the financial balance of the enterprise, that is, the producer surplus should be equal to the fixed cost, we know that R satisfies:

[0085]

[0086] Where F represents the fixed cost of the power company.

[0087] S3. Initialize the boundary based on the load sequence containing load values ​​for each time period in a typical day, update the boundary using the moving boundary method, and divide the time period based on the relationship between the updated boundary and the load values ​​in the load sequence.

[0088] In this embodiment, an improved moving boundary technique is used to divide time periods by minimizing the Davidson-Botting Index (DBI). This method fully considers the intra-class and inter-class distances between different peak and valley time periods and establishes a peak and valley time period division model. By changing the moving variable V fv，m 、V pf，n The optimal peak-valley time division method is solved by the position. The specific steps are as follows:

[0089] S31. Input a typical target first load sequence, sort the first load sequence from low to high, and obtain a second load sequence, where the load sequence includes hourly load values.

[0090] Specifically, the typical daily load sequence L = {l1, l2, ..., l 24}, sort L from low to high, and get a new sequence L′={l′1, l′2,…l′ 24}.

[0091] S32: Initialize a first boundary and a second boundary according to the second load sequence, where the first boundary is smaller than the second boundary.

[0092] Specifically, set the initial value of the decision variable, such as V fv，m =l′1,V pf，n =l′2, where m=T, n=2T, and m∈[T, 24-2T], n∈[2T, 24-T], calculate the objective function value.

[0093] S33 , iteratively updating the first boundary and the second boundary by moving the boundary until n=24-T, and terminating the iteration.

[0094] In this way, the minimum objective function value and its corresponding m and n values ​​are compared and obtained.

[0095] S34. Compare the load value of each hour in the second load sequence with the first boundary and the second boundary of the updated iteration. If the load value is less than or equal to the first boundary, the hour corresponding to the load value is the valley period. If the load value is greater than the first boundary and less than the second boundary, the hour corresponding to the load value is the normal period. If the load value is greater than the second boundary, the hour corresponding to the load value is the peak period.

[0096] Specifically, when l′ t ≤V fv，m , l′ t For valley load; when V fv，m ≤l′ t ≤V pf，n , l′ t It is the load during normal period; otherwise it is the load during peak period.

[0097] In this embodiment, after implementing peak-valley time-of-use electricity pricing, the differences in electricity prices between different time periods will guide residents to change their electricity usage patterns and redistribute load. A user demand response model is established by constructing an electricity price elasticity matrix. The price elasticity of peak-valley time-of-use electricity pricing refers to the percentage change in electricity consumption caused by the percentage change in electricity prices during different time periods. This includes both self-elasticity and cross-elasticity. The effect of electricity price changes within the same time period on electricity consumption changes is self-elasticity, while the effect of electricity price changes in other time periods on electricity consumption changes during the same time period is cross-elasticity.

[0098] Given that residents of different income groups have significant differences in consumption habits and affordability, which are also key factors affecting price demand elasticity, a step-by-step approach is adopted to divide overall residential electricity demand into different market segments. By analyzing the intensity of demand response to peak and valley time-of-use electricity prices among residents in each sub-group market segment based on the dimension of electricity consumption, the elasticity coefficient expression for the i-th tier is:

[0099] Self-elastic coefficient:

[0100] Cross elastic coefficient:

[0101] Where l and n are peak, flat and valley periods respectively.

[0102] Taking the peak, flat and valley periods as an example, the electricity price elasticity matrix of the i-th electricity consumption segment market can be expressed as:

[0103]

[0104] Where, e i,pp 、e i,ff 、ei,vv They represent the self-elasticity coefficients of the peak, flat and valley periods in the i-th electricity consumption segment market, which are negative values. The rest are cross-elasticity coefficients, which are positive values.

[0105] The electricity consumption in a certain period is not only related to the change of electricity price in this period but also to the change of electricity price in other periods. To reflect the relationship between electricity price and electricity consumption, the change of electricity consumption of residents in the i-th tier after the implementation of time-of-use electricity price can be expressed as:

[0106]

[0107] Where p i,p 、p i,f 、p i,v represents the peak, flat and valley electricity prices of the i-th tier, q i,p ,q i,f ,q i,v Represents the electricity consumption in each time period.

[0108] S4. Constraints are set in combination with the time period and the electricity price of each gear, and an objective function is set with the goal of minimizing peak load and peak-to-valley difference. A stepped time-of-use electricity price optimization model is established in combination with the constraints and the objective function, and a stepped time-of-use electricity price scheme is obtained by solving the stepped time-of-use electricity price optimization model.

[0109] Specifically, the tiered electricity price system encourages users to save electricity and use electricity reasonably, but lacks interaction with the load status of the power grid. The core goal of introducing time-of-use electricity prices is to improve the shape of the daily load curve to reduce the peak-valley difference, thereby increasing the residential electricity load rate. To achieve this goal, it is necessary to reduce the peak load of the daily load curve and increase the valley load, while reducing the peak-valley load difference and ensuring that the order of peak-valley load is not reversed. Analysis shows that reducing the peak load and increasing the valley load will inevitably lead to a reduction in the peak-valley difference, that is, any combination of the two aspects can ensure the realization of the other aspect. Therefore, this embodiment selects minimizing peak load and minimizing peak-valley difference as the objective function. At the same time, in order to avoid deviations caused by excessive numerical differences between the two, the load peak value and peak-valley difference under the original electricity price are used as reference values ​​and are normalized:

[0110] S41. Establish an objective function with the goal of minimizing peak load and peak-to-valley difference:

[0111]

[0112]

[0113] In the formula, M1 represents minimizing the daily load peak, M2 represents minimizing the daily load peak-to-valley difference, and L i，max , L i，min , L′ i，max , L′i，min They represent the maximum and minimum daily load values ​​of the i-th electricity consumption segment market before electricity price adjustment and the maximum and minimum daily load values ​​after adjustment respectively.

[0114] The above dual objective function is transformed into a single objective function through the weight coefficient method. In view of the equal importance of the two objectives, the final optimization objective expression is determined by evenly distributing the weights:

[0115]

[0116] S42. Setting constraints on electricity price, time period and quantity of electricity, and establishing a stepped time-of-use electricity price optimization model in combination with the objective function and the constraints.

[0117] 1. Electricity price constraints:

[0118] On the one hand, to ensure the interests of users, the peak, flat and valley electricity prices of each level should meet the constraints of the average electricity price level of that level:

[0119]

[0120] On the other hand, the peak, flat and valley electricity prices in the i-th tier should not be higher than the corresponding period electricity prices in the i+1 tier, and the valley electricity price should not be lower than a certain lower limit p*, that is:

[0121]

[0122] 2. Time constraints:

[0123] To ensure that the peak, flat and valley periods are not inverted, the load in each period should meet the following requirements: i，v ≤l′ i，f ≤l′ i，p

[0124] 3. Power Constraints:

[0125] After the implementation of the tiered time-of-use electricity price, in order to maintain the stability of the user coverage of each market segment of the original tier, the average total electricity consumption per household should not exceed the corresponding tiered electricity consumption:

[0126]

[0127] Where Δl′ i，p , Δl′ i，f , Δl′ i，v are the average loads of each peak, flat and valley period respectively, T p 、T f 、T v is the duration of each period, N i is the number of users in each level, Q i It is the upper limit of each level of power.

[0128] At the same time, according to the electricity price elasticity matrix, the relationship between electricity demand and price can be expressed as:

[0129]

[0130] Where, They are the electricity consumption during the off-peak and off-valley periods of the original electricity price respectively.

[0131] Then, a numerical example is used to verify the effectiveness of this embodiment:

[0132] This analysis uses the current implementation of tiered residential electricity pricing in Fujian as an example. The region's latest residential electricity pricing policy implements three monthly electricity price tiers, with upper limits of 230 kWh and 420 kWh, and prices of 0.4983, 0.5483, and 0.7983 yuan / kWh, respectively. A simulation analysis uses monthly electricity consumption and load data from urban communities and rural households in a particular city in the region for a specific year. The upper and lower limits for monthly electricity consumption are set at 60 kWh and 600 kWh to eliminate outliers. By analyzing the sample data and combining it with demand response data from various provinces and cities across China and academic research experience, we estimate the electricity price elasticity matrix for each sub-group market segment. The specific data is shown in Table 1.

[0133] Table 1 Elasticity matrix of resident market segments

[0134]

[0135] In addition, it is assumed that the marginal cost of electricity supply to users in different market segments (M C ) remains unchanged in the short term, meaning the marginal cost is a constant value of approximately 0.3932 yuan. Furthermore, based on the affordability of residential users in the region and the grid company's need for cost compensation, the lower limit of the off-peak electricity price is set at 0.282 yuan / kWh.

[0136] MATLAB software was used for data simulation. In the tiered electricity price optimization, the electricity consumption was set to three gradients, and the k-means++ algorithm was used to optimize the calculations in MATLAB to obtain the electricity consumption of each level. The electricity consumption limit values ​​of each level are not strictly integer values ​​or narrow interval values. Considering the applicability of the electricity price scheme, integer values ​​within the interval are used as the boundary electricity consumption of each level. The elastic coefficient η of each level was fitted from the data sample. i 、k i Substituting the demand function of each level into the Ramsey pricing index calculation formula, we get R = 0.2426. i Substituting , R into the electricity price calculation formula yields the average electricity price for each tier. All results are shown in Table 2.

[0137] Table 2 Optimization results of step-by-step electricity consumption and electricity prices

[0138]

[0139] In the table, user distribution = number of users whose monthly electricity consumption is distributed in each gear / total users.

[0140] The initial time-of-use electricity rates for residents in this region are: peak period (8:00 AM to 10:00 PM) and off-peak period (10:00 PM to 8:00 AM). This classification method, because it fails to fully consider residents' actual electricity usage habits, is somewhat subjective and inconsistent with actual conditions. Its low penetration rate makes it ineffective in guiding residents to change their electricity usage patterns and reduce the peak-off-peak difference. Therefore, in this example, a typical daily load curve for residents in a city in this region in a certain year is selected and re-classified using an improved moving boundary technique. This adds a time constraint factor to fully consider the distance between and within each class, effectively improving the classification accuracy.

[0141] The division results are as follows Figure 5 As shown: Peak hours: 10:00-12:00, 17:00-21:00; Normal hours: 7:00-10:00, 12:00-17:00, 21:00-23:00; Off-peak hours: 23:00-6:00 the next day.

[0142] Based on the original data and the above analysis results, the parameters of each constraint condition are obtained, and the particle swarm algorithm is used to solve the optimization model. The initial load value and electricity price of each sub-group market segment are input to obtain the optimization results of the ladder time-of-use electricity price, as shown in Table 3.

[0143] Table 3 Optimization results of ladder time-of-use electricity prices

[0144]

[0145] In order to evaluate the implementation effect of the ladder time-of-use electricity price, the maximum load, minimum load, load peak-valley difference (ΔL i，max , ΔL i，min ,Δ(L i，max -L i，min ))The growth rates of the three are used as evaluation indicators to compare the electricity price system with the original electricity price. The comparison results are shown in Table 4.

[0146] Table 4 Analysis of changes in residential load

[0147]

[0148] The load variation index shows an initial upward and then downward trend with the number of tiers. This is because the design of each tier of electricity pricing takes into account the electricity consumption characteristics and behavioral responses of different user groups. First, the first tier of electricity pricing focuses on meeting daily essential electricity needs, resulting in lower load elasticity and limited room for users to adjust their loads. By the second tier, users' basic living needs are already met, and they have greater autonomy in allocating electricity consumption and are more sensitive to price fluctuations. Therefore, they will consciously adjust their electricity consumption to adapt to the peak-valley pricing strategy. The third tier, targeting high-income luxury users, accounts for a relatively small proportion of electricity consumption (only 7.47% of total electricity consumption) and is less sensitive to electricity prices. Therefore, their load adjustments have a smaller impact on the overall load, resulting in the load variation index showing an initial upward and then downward trend. This shows that the optimized tiered time-of-use electricity pricing system effectively achieves load curve adjustment functions such as "peak load reduction" and "peak shaving and valley filling," completing the load shifting function.

[0149] Example 2

[0150] Please refer to Figure 2 A terminal 1 for optimizing a stepped time-of-use electricity price includes a memory 2, a processor 3, and a computer program stored in the memory 2 and executable on the processor 3. When the processor 3 executes the computer program, each step of a method for optimizing a stepped time-of-use electricity price in embodiment 1 is implemented.

[0151] In summary, the present invention provides a method and terminal for optimizing step time-of-use electricity prices, which optimize step and time-of-use electricity prices respectively. In the optimization of step electricity prices, a step electricity quantity optimization model is constructed based on the k-means++ algorithm to make the electricity quantity boundaries of each level more in line with reality, and further use the Ramsey pricing theory to determine the electricity price of each level; in the optimization of time-of-use electricity prices, an improved moving boundary technology is used to fully consider the distance between classes and within classes to improve the accuracy of time period division, and based on the demand response of the electricity price elasticity matrix, a step time-of-use electricity price optimization model is established with the goal of minimizing peak load and peak-valley difference.

[0152] Through case analysis, it was determined that the k-means++ algorithm, based on actual user electricity usage data, fully demonstrates its objectivity and applicability. The application of Ramsey pricing theory to electricity pricing not only effectively protects the social welfare of residential users but also minimizes the negative impact on electricity resource allocation, thereby achieving a fair, efficient, and sustainable development of the electricity market. A stepped time-of-use electricity price optimization model, which aims to minimize peak load and peak-to-valley differences, effectively ensures the smooth implementation of peak-to-valley shifting strategies while not affecting residents' daily electricity needs. This guides residents to use electricity rationally and achieves a balanced distribution of electricity load.

[0153] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's description and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for optimizing a ladder time-of-use electricity price, characterized in that: Including steps: S1. Determine the number of electricity price levels and the corresponding electricity volume for each level through a clustering algorithm, including: S11. Obtain a sample set of power data, calculate the number of clusters using the elbow method, randomly select a sample from the sample set as the first cluster center, and use the remaining samples farthest from the first cluster center as new second cluster centers, until the sum of the number of the first cluster centers and the second cluster centers reaches the specified number of clusters; S12, calculating the distances of the remaining samples to the first cluster center or the second cluster center in sequence, and classifying the remaining samples into the cluster corresponding to the cluster center with the shortest distance; S13, recalculating the center point of each cluster; S14, repeating steps S12 and S13 until the center point of each cluster no longer changes, outputting a clustering result, determining the number of electricity price levels according to the number of clusters in the clustering result, and determining the electricity quantity boundary value corresponding to each level according to the cluster boundary of the clustering result; S2. Pricing electricity prices for each tier based on Ramsey pricing, including: Set a preset electricity price for the first tier of electricity prices; Pricing for electricity prices in tier 2 and above based on Ramsey pricing: Where, p i Indicates the i Average electricity price per tier, M c represents the long-run marginal cost, η i Indicates the i Level demand response intensity, R represents the Ramsey index; S3. Initializing the boundary based on a load sequence containing load values ​​for each time period in a typical day, updating the boundary using a moving boundary method, and dividing the time period based on the relationship between the updated boundary and the load values ​​in the load sequence, including: S31. Input a first load sequence of a typical day, sort the first load sequence from low to high, and obtain a second load sequence, where the load sequence includes hourly load values; S32: Initialize a first boundary and a second boundary according to the second load sequence, wherein the first boundary is smaller than the second boundary; S33, iteratively updating the first boundary and the second boundary by moving the boundary until the iteration result meets a preset condition; S34. Compare the load value of each hour in the second load sequence with the first boundary and the second boundary of the updated iteration. If the load value is less than or equal to the first boundary, the hour corresponding to the load value is a valley period. If the load value is greater than the first boundary and less than the second boundary, the hour corresponding to the load value is a normal period. If the load value is greater than the second boundary, the hour corresponding to the load value is a peak period. S4. Constraints are set in combination with the time period and the electricity price of each gear, and an objective function is set with the goal of minimizing peak load and peak-to-valley difference. A stepped time-of-use electricity price optimization model is established in combination with the constraints and the objective function, and a stepped time-of-use electricity price scheme is obtained by solving the stepped time-of-use electricity price optimization model.

2. The method for optimizing a ladder time-of-use electricity price according to claim 1, characterized in that: Step S4 includes: The objective function is established with the goal of minimizing peak load and minimizing peak-to-valley difference: Where, M 1 means minimizing the daily load peak, M 2 means minimizing the daily load peak-valley difference, 、 、 、 Respectively represent the first i The maximum and minimum daily load values ​​and the adjusted maximum and minimum daily load values ​​for each electricity consumption segment market; The constraints of electricity price, time period and quantity are set, and a step-by-step time-of-use electricity price optimization model is established in combination with the objective function and the constraints.

3. An optimization terminal for ladder time-of-use electricity prices, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the following steps are implemented: S1. Determine the number of electricity price levels and the corresponding electricity volume for each level through a clustering algorithm, including: S11. Obtain a sample set of power data, calculate the number of clusters using the elbow method, randomly select a sample from the sample set as the first cluster center, and use the remaining samples farthest from the first cluster center as new second cluster centers, until the sum of the number of the first cluster centers and the second cluster centers reaches the specified number of clusters; S12, calculating the distances of the remaining samples to the first cluster center or the second cluster center in sequence, and classifying the remaining samples into the cluster corresponding to the cluster center with the shortest distance; S13, recalculating the center point of each cluster; S14, repeating steps S12 and S13 until the center point of each cluster no longer changes, outputting a clustering result, determining the number of electricity price levels according to the number of clusters in the clustering result, and determining the electricity quantity boundary value corresponding to each level according to the cluster boundary of the clustering result; S2. Pricing electricity prices for each tier based on Ramsey pricing, including: Set a preset electricity price for the first tier of electricity prices; Pricing for electricity prices in tier 2 and above based on Ramsey pricing: Where, p i Indicates the i Average electricity price per tier, M c represents the long-run marginal cost, η i Indicates the i Level demand response intensity, R represents the Ramsey index; S3. Initializing the boundary based on a load sequence containing load values ​​for each time period in a typical day, updating the boundary using a moving boundary method, and dividing the time period based on the relationship between the updated boundary and the load values ​​in the load sequence, including: S31. Input a first load sequence of a typical day, sort the first load sequence from low to high, and obtain a second load sequence, where the load sequence includes hourly load values; S32: Initialize a first boundary and a second boundary according to the second load sequence, wherein the first boundary is smaller than the second boundary; S33, iteratively updating the first boundary and the second boundary by moving the boundary until the iteration result meets a preset condition; S34. Compare the load value of each hour in the second load sequence with the first boundary and the second boundary of the updated iteration. If the load value is less than or equal to the first boundary, the hour corresponding to the load value is a valley period. If the load value is greater than the first boundary and less than the second boundary, the hour corresponding to the load value is a normal period. If the load value is greater than the second boundary, the hour corresponding to the load value is a peak period. S4. Constraints are set in combination with the time period and the electricity price of each gear, and an objective function is set with the goal of minimizing peak load and peak-to-valley difference. A stepped time-of-use electricity price optimization model is established in combination with the constraints and the objective function, and a stepped time-of-use electricity price scheme is obtained by solving the stepped time-of-use electricity price optimization model.

4. The optimization terminal for step-by-step time-of-use electricity prices according to claim 3, characterized in that: Step S4 includes: The objective function is established with the goal of minimizing peak load and minimizing peak-to-valley difference: Where, M 1 means minimizing the daily load peak, M 2 means minimizing the daily load peak-valley difference, 、 、 、 Respectively represent the first i The maximum and minimum daily load values ​​and the adjusted maximum and minimum daily load values ​​for each electricity consumption segment market; The constraints of electricity price, time period and quantity are set, and a step-by-step time-of-use electricity price optimization model is established in combination with the objective function and the constraints.

Images