Multi-source data fusion adaptive segmented power market bidding method and system

Abstract

The application discloses a kind of multi-source data fusion adaptive segmented power market declaration method and system, it is related to electric power market transaction technical field, including step S1: multi-source data acquisition and fusion.Compared with prior art, the beneficial effects of the present application are: the present application fuses three major data sources of medium and long term contract, load prediction, electricity price prediction, and quantifies the importance of time period through a weighted urgency model, reduces overall bias, increases revenue, a dynamic boundary generation mechanism automatically adjusts the range of declaration volume according to the actual bias pattern, the weighted coverage is high in mixed scenarios, and the adjustment effect is directly mapped to the bidding strategy with the comprehensive score as the bridge, the proportion of successful bidding electricity is increased, the average successful bidding price is better, the multi-dimensional urgency model guides to give priority to guarantee high urgency period, when the adjustment capacity is insufficient, the demand satisfaction rate of high urgency period is improved, and the integrated software system can realize one-key generation of the next day declaration scheme, improve the operation efficiency and decision-making intelligent level.

Description

Technical Field

[0001] This invention relates to the field of electricity market trading technology, specifically a multi-source data fusion adaptive segmented electricity market declaration method and system. Background Technology

[0002] With the large-scale grid connection of new energy sources and the deepening of power market reforms, new market players such as virtual power plants and load aggregators need to participate in the spot market to obtain profits and balance risks. Currently, the domestic electricity spot market generally adopts a "volume-based bidding" method, requiring market participants to submit multiple volume-price curves (such as a five-segment curve), which are then cleared uniformly by the market clearing agency. The quality of the bidding strategy directly affects the market participants' profits and risk control capabilities.

[0003] The existing application strategies mainly face the following technical challenges: data fragmentation and single decision-making basis: existing methods usually rely on a single data source, such as bidding based solely on historical electricity price curves or simple load forecasts. They fail to effectively integrate multi-source heterogeneous data such as the breakdown of medium- and long-term contract electricity volume, day-ahead load forecasts, and day-ahead electricity price forecasts, resulting in a disconnect between the application volume and actual demand.

[0004] The rigidity and poor adaptability of the application curve: Traditional segmented applications (such as three or five-segment quotations) usually use fixed segment points (such as equal intervals or fixed proportions), which fail to dynamically adjust according to the direction, size and urgency of the actual daily adjustment needs. In scenarios with mixed needs (both upward and downward adjustments), fixed segments may not be able to meet both demands at the same time, resulting in a waste of resources.

[0005] Disconnect between quantity and price decisions: In existing methods, the determination of the declared quantity and the declared price are usually separate, failing to take the improvement effect (i.e. "value") of different declaration segments on the overall deviation of the system as the core basis for price determination, resulting in a lack of economic rationality in the quotations.

[0006] Insufficient dynamic response to regulation capacity and deviation: Existing methods consider the constraints of regulation resources in a static way and do not quantitatively assess the "urgency" of deviation in each regulation period. When regulation demand exceeds regulation capacity, there is a lack of intelligent allocation mechanism.

[0007] Therefore, there is an urgent need for a power market application method that can deeply integrate multi-source data, adaptively generate application curves, and achieve intelligent decision-making based on quantity and price coordination. Summary of the Invention

[0008] This invention provides a multi-source data fusion adaptive segmented power market application method and system to solve the problems mentioned in the background art.

[0009] According to an embodiment of the present invention, in a first aspect, a multi-source data fusion adaptive segmented power market application method is provided, comprising the following steps:

[0010] Step S1: Multi-source data acquisition and fusion, acquiring the medium- and long-term contract breakdown electricity volume for each time period of the target day as the basic electricity volume sequence, the load forecast value as the predicted load sequence, the spot market clearing price forecast value as the predicted electricity price sequence, as well as the regulation capacity parameters and the target regulation period set;

[0011] Step S2: Time-based deviation and urgency analysis. For each time period in the target adjustment time period set, calculate the power consumption deviation, and calculate a multi-dimensional urgency score based on the power consumption deviation, the basic power consumption sequence, the predicted electricity price sequence, whether the time period belongs to the peak period, and whether there is an adjustment demand.

[0012] Step S3: Adaptive segmented reporting quantity design. Based on the sign distribution of the power deviation of all target adjustment periods, the deviation pattern is identified, and the upper and lower limits of the reporting quantity are dynamically determined by enumerating candidate boundaries and calculating the urgency-weighted coverage. Then, a set of cumulative adjustment quantities containing multiple reporting segments is generated within the upper and lower limits.

[0013] Step S4: Performance evaluation of the application segment. Simulate the application of each application segment, calculate the average remaining absolute deviation of each application segment for all target adjustment periods, calculate the deviation improvement score based on the average remaining absolute deviation, and calculate the average urgency score for all target adjustment periods. Weight and fuse the deviation improvement score and the average urgency score to obtain the comprehensive score of each application segment.

[0014] Step S5: Value-based intelligent pricing. After normalizing the comprehensive score, the average of the predicted electricity price series is used as the benchmark price. Different pricing formulas are used to generate initial prices for different declaration segments. When the average urgency score of all target adjustment periods exceeds the preset threshold, the prices of the upward and downward adjustment segments are multiplied by different adjustment coefficients respectively. Finally, monotonicity correction is performed to obtain the final declaration price of each declaration segment.

[0015] Step S6: Generate the final application plan and output the application plan containing the cumulative adjustment amount, semantic tags, final application price and performance analysis data of each application segment. The performance analysis data includes at least one of the following: average residual deviation, improvement score and comprehensive score.

[0016] As a further aspect of the present invention: In step S1, the basic electricity consumption sequence is obtained by querying the contract electricity consumption table in the database, summarizing them according to contract type, and then aggregating the time series data into hourly electricity consumption and weighting the average; the predicted electricity price sequence is obtained by querying the electricity price prediction table in the database, averaging it by hour, and using robust statistical methods to remove outliers; the predicted load sequence is obtained by querying the predicted electricity consumption meter in the database, aggregating it by hour, and removing outliers; the regulation capacity parameter can be input by the user or obtained based on historical regulation capacity statistics; the target regulation time period set is a number of hourly time periods specified by the user or determined by the strategy that require declaration and regulation.

[0017] As a further aspect of the present invention: in step S2, the formula for calculating the multidimensional urgency score Urgency(t) is as follows:

[0018] Urgency(t) = min(deviation ratio factor + absolute deviation factor + electricity price level factor + peak period factor + adjustment demand factor, 100);

[0019] Wherein, the deviation scaling factor = min(|Deviation(t)| / |Q_base(t)|×100,100);

[0020] Absolute deviation factor = min(|Deviation(t)| / 5×50,50);

[0021] Electricity price level factor = min(P_spot(t) / 20,30);

[0022] The peak period factor is set to 20 when the period falls within the system load peak, and 0 otherwise.

[0023] The adjustment demand factor is set to 10 when Deviation(t) ≠ 0, and 0 otherwise.

[0024] Deviation(t) represents the charge deviation, and the formula for calculating the charge deviation is:

[0025] Deviation(t)=Q_pred(t)-Q_base(t);

[0026] Q_pred(t) represents the user's expected electricity consumption, Q_base(t) represents the base electricity consumption, and P_spot(t) represents the predicted electricity price.

[0027] As a further aspect of the present invention, step S3 specifically includes:

[0028] Step S301: Deviation mode identification. If all power deviations are ≥0, it is a full increase mode; if all power deviations are ≤0, it is a full decrease mode; otherwise, it is a mixed mode.

[0029] Step S302: Calculate net adjustment demand, including total upward adjustment demand and total downward adjustment demand;

[0030] Step S303: Dynamic boundary generation. Based on the identified pattern, determine the lower limit L and upper limit U of the declared amount. For the full upward or full downward adjustment pattern, the boundary is determined based on the minimum deviation, maximum deviation and adjustment capability parameters. For the mixed pattern, by enumerating multiple candidate boundary combinations, the urgency weighted coverage under each combination is calculated, and the boundary combination that maximizes the coverage is selected as the final (L, U).

[0031] Step S304: Segment quantity generation and labeling. Within the defined boundary (L, U), five monotonically increasing declaration quantities are generated through linear interpolation, and semantic labels are assigned to each segment based on whether the boundary crosses the zero point.

[0032] As a further aspect of the present invention: in step S303, the formula for calculating the urgency-weighted coverage is as follows:

[0033] coverage=Σ_t[w(t)×coverage_rate(t)] / Σ_tw(t);

[0034] Where w(t) = Urgency(t) / 100, and coverage_rate(t) is the proportion of the deviation in time period t that is covered by the declaration volume of this group;

[0035] If Deviation(t) > 0 and A[i] ≥ Deviation(t), then the coverage rate is 1; otherwise, take max(A) / Deviation(t).

[0036] If Deviation(t) < 0 and A[i] ≤ Deviation(t), then the coverage rate is 1; otherwise, take min(A) / Deviation(t).

[0037] If Deviation(t) = 0: the coverage rate is 1;

[0038] Where A[i] is the candidate application segment, and min(A) is the maximum candidate application segment.

[0039] As a further aspect of the present invention: step S4 specifically includes:

[0040] Step S401: Simulation application: For each declared segment and each adjustment period, calculate the actual adjustment amount and remaining deviation under the adjustment capacity limit according to the direction and size of the declared segment;

[0041] Step S402: Calculate the mean residual absolute deviation, mean residual absolute deviation = (1 / |T|)Σ_t |remaining(t)|;

[0042] Where |T| is the total number of target adjustment periods, and |remaining(t)| is the absolute value of the remaining deviation that still exists in the t-th period after adjustment of a certain reporting segment;

[0043] Step S403: Calculate the deviation improvement score, deviation improvement score = max(0, 100 - avg_abs_remaining × 10);

[0044] Where avg_abs_remaining is the average residual absolute deviation;

[0045] Step S404: Calculate the overall score, Overall Score = Deviation Improvement Score × (1 + Urgency Weight × α);

[0046] Wherein, the urgency weight = average urgency / 100, average urgency = (1 / |T|)Σ_tUrgency(t), and α is an adjustable coefficient between 0.3 and 0.7.

[0047] As a further aspect of the present invention: step S5 specifically includes:

[0048] Step 501: Normalization process, normalize the comprehensive scores of all application segments to the [0,1] interval;

[0049] Step 502: Benchmark price anchoring, using the average of the predicted electricity prices for all target adjustment periods as the market average price;

[0050] Step 503: Value-driven pricing. The temporary price for the upward adjustment phase = (market average price × first benchmark coefficient) × (first basic competitiveness factor + first adjustment coefficient × (1 - normalized score)). The temporary price for the downward adjustment phase = (market average price × second benchmark coefficient) × (second basic competitiveness factor - second adjustment coefficient × normalized score).

[0051] Step 504: Monotonicity assurance and minimum interval correction. Force correction is applied to the generated price series to ensure that the bid price strictly decreases as the bid volume increases, and the price difference between adjacent segments is not less than 1% of the highest bid price.

[0052] Secondly, this invention proposes a multi-source data fusion adaptive segmented electricity market declaration system, the system comprising:

[0053] The data acquisition module is used to obtain the medium- and long-term contract electricity volume, forecasted electricity consumption, medium- and long-term average holding price, and forecasted spot price from the database for each time period of the target day.

[0054] The deviation and urgency acquisition module is connected to the data acquisition module and is used to calculate the power deviation and multi-dimensional urgency score for each time period based on the data provided by the data acquisition module.

[0055] An adaptive segmentation module, connected to the deviation and urgency acquisition module, is used to identify deviation patterns based on the sign distribution of the power deviation, and dynamically generate a set of cumulative adjustment amounts containing multiple reporting segments based on the deviation patterns and the multidimensional urgency scores.

[0056] The performance evaluation module, connected to the adaptive segmentation module, is used to simulate each application segment and calculate the average residual absolute deviation, deviation improvement score, and overall score for each application segment.

[0057] The intelligent pricing module, connected to the performance evaluation module, normalizes the comprehensive score and uses the average of the predicted electricity price sequence as the benchmark price. It then uses different pricing formulas to generate initial prices for different declaration segments and performs monotonicity correction to obtain the final declaration price for each declaration segment.

[0058] The output module, connected to the intelligent pricing module, is used to output the declaration scheme, which includes the cumulative adjustment amount, semantic tags, final declaration price, and performance analysis data for each declaration segment.

[0059] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention provides a multi-source data fusion adaptive segmented power market application method and system that integrates three major data sources: medium- and long-term contracts, load forecasts, and electricity price forecasts. By quantifying the importance of time periods through a weighted urgency model, it can reduce the overall deviation by more than 30% and increase revenue by 15%-25%. The dynamic boundary generation mechanism automatically adjusts the application volume range according to the actual deviation pattern. The weighted coverage in mixed scenarios can reach more than 90%, with strong adaptability and high coverage. Using "comprehensive score" as a bridge, the adjustment effect is directly mapped to the bidding strategy, increasing the winning bid volume ratio by 20% and the average winning bid price. The multi-dimensional urgency model guides priority to ensure high-urgency periods. When the adjustment capacity is insufficient, the demand satisfaction rate of high-urgency periods is increased by more than 40%. Resources are optimized, and the system is integrated into a complete software system that can generate the next day's application plan with one click, improving operational efficiency and the level of intelligent decision-making. Attached Figure Description

[0060] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the 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.

[0061] Figure 1This is a flowchart of a multi-source data fusion adaptive segmented power market application method provided in an embodiment of the present invention.

[0062] Figure 2 This is a first flowchart for the adaptive segmented reporting volume design provided in an embodiment of the present invention.

[0063] Figure 3 A flowchart for performance evaluation of the application segment provided in an embodiment of the present invention.

[0064] Figure 4 This is a first flowchart of value-based smart pricing provided for an embodiment of the present invention.

[0065] Figure 5 The second flowchart for the adaptive segmented declaration volume design provided in the embodiments of the present invention.

[0066] Figure 6 This is a second flowchart of value-based smart pricing provided for an embodiment of the present invention.

[0067] Figure 7 This is a block diagram of a multi-source data fusion adaptive segmented power market declaration system provided in an embodiment of the present invention. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Please see Figure 1 A multi-source data fusion adaptive segmented electricity market declaration method includes the following steps:

[0070] Step S1: Multi-source data acquisition and fusion, acquiring the medium- and long-term contract breakdown electricity volume for each time period of the target day as the basic electricity volume sequence, the load forecast value as the predicted load sequence, the spot market clearing price forecast value as the predicted electricity price sequence, as well as the regulation capacity parameters and the target regulation period set;

[0071] Step S2: Time-based deviation and urgency analysis. For each time period in the target adjustment time period set, calculate the power consumption deviation, and calculate a multi-dimensional urgency score based on the power consumption deviation, the basic power consumption sequence, the predicted electricity price sequence, whether the time period belongs to the peak period, and whether there is an adjustment demand.

[0072] Step S3: Adaptive segmented reporting quantity design. Based on the sign distribution of the power deviation of all target adjustment periods, the deviation pattern is identified, and the upper and lower limits of the reporting quantity are dynamically determined by enumerating candidate boundaries and calculating the urgency-weighted coverage. Then, a set of cumulative adjustment quantities containing multiple reporting segments is generated within the upper and lower limits.

[0073] Step S4: Performance evaluation of the application segment. Simulate the application of each application segment, calculate the average remaining absolute deviation of each application segment for all target adjustment periods, calculate the deviation improvement score based on the average remaining absolute deviation, and calculate the average urgency score for all target adjustment periods. Weight and fuse the deviation improvement score and the average urgency score to obtain the comprehensive score of each application segment.

[0074] Step S5: Value-based intelligent pricing. After normalizing the comprehensive score, the average of the predicted electricity price series is used as the benchmark price. Different pricing formulas are used to generate initial prices for different declaration segments. When the average urgency score of all target adjustment periods exceeds the preset threshold, the prices of the upward and downward adjustment segments are multiplied by different adjustment coefficients respectively. Finally, monotonicity correction is performed to obtain the final declaration price of each declaration segment.

[0075] Step S6: Generate the final application plan and output the application plan containing the cumulative adjustment amount, semantic tags, final application price and performance analysis data of each application segment. The performance analysis data includes at least one of the following: average residual deviation, improvement score and comprehensive score.

[0076] Further, in step S1, the basic electricity consumption sequence is obtained by querying the contract electricity consumption table in the database, summarizing them according to contract type, and then weighting the time series data into hourly electricity consumption; the predicted electricity price sequence is obtained by querying the electricity price prediction table in the database, averaging it by hour, and using robust statistical methods to remove outliers; the predicted load sequence is obtained by querying the predicted electricity consumption meter in the database, aggregating it by hour, and removing outliers; the regulation capacity parameter can be input by the user or obtained based on historical regulation capacity statistics; the target regulation time period set is a number of hourly time periods specified by the user or determined by the strategy that require declaration and regulation.

[0077] Furthermore, in step S2, the formula for calculating the multidimensional urgency score Urgency(t) is as follows:

[0078] Urgency(t) = min(deviation ratio factor + absolute deviation factor + electricity price level factor + peak period factor + adjustment demand factor, 100) (1);

[0079] Wherein, the deviation scaling factor = min(|Deviation(t)| / |Q_base(t)|×100,100) (2;

[0080] Absolute deviation factor = min(|Deviation(t)| / 5×50,50) (3;

[0081] Electricity price level factor = min(P_spot(t) / 20,30) (4);

[0082] The peak period factor is set to 20 when the period falls within the system load peak, and 0 otherwise.

[0083] The adjustment demand factor is set to 10 when Deviation(t) ≠ 0, and 0 otherwise.

[0084] Deviation(t) represents the charge deviation, and the formula for calculating the charge deviation is:

[0085] Deviation(t)=Q_pred(t)-Q_base(t) (5);

[0086] Q_pred(t) represents the user's expected electricity consumption, Q_base(t) represents the base electricity consumption, and P_spot(t) represents the predicted electricity price.

[0087] In a specific embodiment, the weight range of the above-mentioned deviation ratio factor is 0-40, the weight range of the absolute deviation factor is 0-25, the weight range of the electricity price level factor is 0-15, the weight range of the peak period factor is 0-10, and the weight range of the adjustment demand factor is 0-10.

[0088] In this embodiment, please refer to Figure 2 and Figure 5 , Figure 5 The second flowchart designed for adaptive segmented reporting volume illustrates the internal sub-processes in steps S301 to S304, including deviation pattern recognition, net demand calculation, candidate boundary enumeration, coverage calculation, and boundary selection. Step S3 specifically includes:

[0089] Step S301: Deviation mode identification. If all power deviations are ≥0, it is a full increase mode; if all power deviations are ≤0, it is a full decrease mode; otherwise, it is a mixed mode.

[0090] Step S302: Calculate net adjustment demand, including total upward adjustment demand and total downward adjustment demand;

[0091] Step S303: Dynamic boundary generation. Based on the identified pattern, determine the lower limit L and upper limit U of the declared amount. For the full upward or full downward adjustment pattern, the boundary is determined based on the minimum deviation, maximum deviation and adjustment capability parameters. For the mixed pattern, by enumerating multiple candidate boundary combinations, the urgency weighted coverage under each combination is calculated, and the boundary combination that maximizes the coverage is selected as the final (L, U).

[0092] Step S304: Segment quantity generation and labeling. Within the defined boundary (L, U), five monotonically increasing declaration quantities are generated through linear interpolation, and semantic labels are assigned to each segment based on whether the boundary crosses the zero point.

[0093] Furthermore, in step S303, the formula for calculating the urgency-weighted coverage is as follows:

[0094] coverage=Σ_t[w(t)×coverage_rate(t)] / Σ_tw(t) (6);

[0095] Where w(t) = Urgency(t) / 100, and coverage_rate(t) is the proportion of the deviation in time period t that is covered by the declaration volume of this group;

[0096] If Deviation(t) > 0 and A[i] ≥ Deviation(t), then the coverage rate is 1; otherwise, take max(A) / Deviation(t).

[0097] If Deviation(t) < 0 and A[i] ≤ Deviation(t), then the coverage rate is 1; otherwise, take min(A) / Deviation(t).

[0098] If Deviation(t) = 0: the coverage rate is 1;

[0099] Where A[i] is the candidate application segment, and min(A) is the maximum candidate application segment.

[0100] In a specific embodiment, the numerical range of the above full upward regulation mode: set the lower regulation boundary as the minimum deviation in the regulation period, that is, L = the minimum value among the electric quantity deviations, and take the minimum value of the maximum deviation and the maximum upward regulation capability as the upper boundary, that is, U = min(maximum deviation, maximum upward regulation capability). If the difference of U-L is less than 1.0, U is forcibly widened to L + maximum upward regulation capability × 0.5; the numerical range of the above full downward regulation mode: set the lower regulation boundary as the maximum value of the maximum deviation and the maximum downward regulation capability, that is, L = max(maximum deviation, -maximum downward regulation capability), and the upper boundary is the minimum deviation in the regulation period, that is, U = the minimum value among the electric quantity deviations. If the difference of U-L is greater than -1.0, that is, the interval length is less than 1, L is widened leftward to U - maximum downward regulation capability × 0.5; the above hybrid mode performs enumeration by trial-selecting candidate boundaries, generates five-segment values for each group of candidate boundaries, and selects the candidate with the highest coverage ratio in the regulation period, that is, selects the boundary combination (L, U) that maximizes the "urgency-weighted coverage"; the above total upward regulation requirement = Σmax(Deviation(t), 0), total downward regulation requirement = Σ|min(Deviation(t), 0)|, net regulation requirement = total upward regulation requirement - total downward regulation requirement; the above candidate boundary combinations include actual deviation range, symmetric range, near-maximum capability range and extended range; the declared quantity in the above step S304 = L + (U-L)*(i / (N-1)), where N=5, i=0, 1, 2, 3, 4.

[0101] According to whether the boundary value crosses zero, five-segment values and semantic tags are generated, and semantic tags are assigned to each segment:

[0102] If L < 0 < U, the tags are: strong downward regulation, medium downward regulation, zero regulation, medium upward regulation, strong upward regulation;

[0103] If L ≥ 0, the tags are: weakest upward regulation, light upward regulation, medium upward regulation, strong upward regulation, strongest upward regulation;

[0104] If U ≤ 0, the tags are: strongest downward regulation, strong downward regulation, medium downward regulation, light downward regulation, weakest downward regulation.

[0105] Finally output five-segment recommended declared quantities.

[0106] Please refer to Figure 3 , step S4 specifically includes:

[0107] Step S401: simulation application, for each declared segment and each regulation period, calculate the actual regulation quantity and remaining deviation under the limitation of regulation capability according to the direction and magnitude of the declared segment;

[0108] Step S402: calculate the average remaining absolute deviation, average remaining absolute deviation=(1 / |T|)Σ_t|remaining(t)| (7);

[0109] Where |T| is the total number of target adjustment periods, and |remaining(t)| is the absolute value of the remaining deviation that still exists in the t-th period after adjustment of a certain reporting segment;

[0110] Step S403: Calculate the deviation improvement score, deviation improvement score = max(0, 100 - avg_abs_remaining × 10) (8);

[0111] Where avg_abs_remaining is the average residual absolute deviation;

[0112] Step S404: Calculate the overall score, overall score = deviation improvement score × (1 + urgency weight × α) (9);

[0113] Wherein, the urgency weight = average urgency / 100, average urgency = (1 / |T|)Σ_tUrgency(t), and α is an adjustable coefficient between 0.3 and 0.7.

[0114] In this embodiment, please refer to Figure 4 and Figure 6 , Figure 6 The second flowchart for value-based intelligent pricing illustrates the pricing formulas for the upward and downward adjustment phases, the normalized score mapping, and the monotonicity correction process. Inputs include the comprehensive score for each phase, the average holding price of medium- and long-term contracts, and urgency weighting information. Step S5 specifically includes:

[0115] Step S501: Normalization process, normalize the comprehensive scores of all application segments to the [0,1] interval;

[0116] Step S502: Benchmark price anchoring, using the average of the predicted electricity prices for all target adjustment periods as the market average price;

[0117] Step S503: Value-driven pricing, the temporary price for the upward adjustment phase = (market average price × first benchmark coefficient) × (first basic competitiveness factor + first adjustment coefficient × (1 - normalized score)), the temporary price for the downward adjustment phase = (market average price × second benchmark coefficient) × (second basic competitiveness factor - second adjustment coefficient × normalized score) (10);

[0118] Step S504: Monotonicity assurance and minimum interval correction: Force correction is applied to the generated price series to ensure that the bid price strictly decreases as the bid volume increases, and the price difference between adjacent segments is not less than 1% of the highest bid price.

[0119] In a specific embodiment, the above-mentioned average market price × first benchmark coefficient = benchmark price for the upward adjustment segment, average market price × second benchmark coefficient = benchmark price for the downward adjustment segment, first basic competitiveness factor + first adjustment coefficient × (1 - normalized score) is the competitiveness for the upward adjustment segment, second basic competitiveness factor - second adjustment coefficient × normalized score is the competitiveness for the downward adjustment segment, and the benchmark price for the no-adjustment segment (zero-adjustment segment) = average market price, competitiveness = 1.0, and temporary price = benchmark price;

[0120] In the upward adjustment phase, the first benchmark coefficient ranges from 0.8 to 0.9, with a typical value of 0.85. That is, the average market price × 0.85 = the benchmark price of the upward adjustment phase. The first basic competitiveness factor ranges from 0.6 to 0.8, the first adjustment coefficient ranges from 0.2 to 0.4, and the competitiveness of the upward adjustment phase can be 0.7 + 0.3 × (1 - normalized score). Temporary price = benchmark price × competitiveness.

[0121] In the downward adjustment phase, the second benchmark coefficient ranges from 1.1 to 1.2, with a typical value of 1.15. That is, the average market price × 1.15 = the benchmark price in the upward adjustment phase. The second basic competitiveness factor ranges from 0.9 to 1.0, and the second adjustment coefficient ranges from 0.15 to 0.25. The competitiveness in the downward adjustment phase can be 1.0 - 0.2 × normalized score. Temporary price = benchmark price × competitiveness.

[0122] If the urgency weight is greater than 0.7, that is, the average urgency weight is greater than 0.7, the temporary price of the upward adjustment segment is multiplied by 0.95, the temporary price of the downward adjustment segment is multiplied by 0.98, and the temporary price of the zero adjustment segment remains unchanged. If the urgency weight is less than 0.7, the temporary prices of the upward adjustment segment, the downward adjustment segment, and the zero adjustment segment remain unchanged. Finally, all segment prices are collected, monotonicity correction and minimum interval check are performed, and the final price sequence is output.

[0123] Experimental Verification: This embodiment takes the example of a virtual power plant operator submitting a five-segment price curve for the four adjustment periods from 11:00 to 14:00 on January 12, 2026. In step S1, the system connects to the configured MySQL database and executes a query to obtain: the basic power generation sequence Q_base=[15.2,16.8,14.5,17.0]MW (corresponding to 11:00-14:00); the predicted load sequence Q_pred=[18.5,12.3,16.0,20.2]MW; and the predicted electricity price sequence P_spot. =[520.3,510.8,498.7,535.4] yuan / MWh; Regulation capacity: max_up_capacity=19MW, max_down_capacity=7MW; Target regulation time period: [11,12,13,14], execute step S2, calculate the deviation for each time period: +3.3MW at 11:00, -4.5MW at 12:00, +1.5MW at 13:00, +3.2MW at 14:00, take 14:00 as an example to calculate the urgency score: the deviation proportional factor is 18.8, absolute... The deviation factor is set to 32, the electricity price level factor to 26.77, the peak period factor to 20, and the adjustment demand factor to 10, totaling 107.57. The upper limit is set to 100 points. The urgency levels for each time period are: 11:85, 12:95, 13:70, and 14:100. Step S3 is executed. The deviation mode is a mixed mode, with a net upward demand adjustment of 3.5MW. By enumerating candidate boundaries and calculating weighted coverage, the final boundary is selected as L=-2.5, U=8.0, generating five segments: [-2.5, -0.875, 0.75, 2.37]. [5,4.0], tagged with ['Strong Downward Adjustment', 'Medium Downward Adjustment', 'Zero Adjustment', 'Medium Upward Adjustment', 'Strong Upward Adjustment'], execute step S4, taking the "Medium Downward Adjustment" segment (-0.875MW) as an example, the average remaining absolute deviation is calculated to be 2.906MW, the improvement score is 70.94, the average urgency is 87.5, and the comprehensive score is 102.0. The scores for each segment are as follows: Strong Downward Adjustment 111.4, Medium Downward Adjustment 102.0, Zero Adjustment 109.2, Medium Upward Adjustment 118.2, Strong Upward Adjustment 127.1. The comprehensive scores for each segment are calculated similarly as shown in Table 1:

[0124] part Quantity (MW) Mean residual deviation Improvement score Overall Score strong downward adjustment -2.5 2.25 77.5 111.4 Medium to low -0.875 2.906 70.94 102.0 Zero adjustment 0.75 2.406 75.94 109.2 Medium to high 2.375 1.781 82.19 118.2 Strong upward adjustment 4.0 1.156 88.44 127.1

[0125] Table 1

[0126] Execute step S5, normalized score: minimum 102.0, maximum 127.1, resulting in normalized scores for each segment: [0.31, 0, 0.30, 0.62, 1.00]. Market average price: 516.3 yuan / MWh; benchmark price for the upward adjustment segment: 438.9 yuan / MWh; benchmark price for the downward adjustment segment: 593.7 yuan / MWh; pricing: [545.9 (strong downward adjustment), 581.8 (medium downward adjustment), 516.3]. [(Zero adjustment), 339.4 (medium to high adjustment), 291.8 (strong to high adjustment)], Monotonicity correction: sorted by quantity (already increasing): -2.5, -0.875, 0.75, 2.375, 4.0; Original price: 545.9, 581.8, 516.3, 339.4, 291.8, forced decrease required: 581.8 > 545.9, therefore 581.8 is changed to 545.9 × 0 0.99 = 540.4; 516.3 < 540.4, satisfying the decreasing condition; 339.4 < 516.3, satisfying the condition; 291.8 < 339.4, satisfying the condition. Minimum interval: highest price 545.9 × 1% = 5.46. Check the differences between adjacent prices: 540.4 - 516.3 = 24.1 > 5.46, 516.3 - 339.4 = 176.9 > 5.46, 339.4 - 291.8 = 47.6 > 5.46, no adjustment is needed. After adjustment by competitiveness factor, urgency and monotonicity, the temporary price sequence is: [545.9 (strong downward adjustment), 540.4 (medium downward adjustment), 516.3 (zero adjustment), 339.4 (medium upward adjustment), 291.8 (strong upward adjustment)]. The price difference between adjacent segments meets the minimum interval requirement. Execute step S6 and output the declaration scheme table, as shown in Table 2 below:

[0127] Section Label Application volume (MW) Price quote (RMB / MWh) Mean residual deviation (MW) Improvement score Overall Score 1 strong downward adjustment -2.5 545.9 2.25 77.5 111.4 2 Medium to low -0.875 540.4 2.906 70.94 102.0 3 Zero adjustment 0.75 516.3 2.406 75.94 109.2 4 Medium to high 2.375 339.4 1.781 82.19 118.2 5 Strong upward adjustment 4.0 291.8 1.156 88.44 127.1

[0128] Table 2

[0129] In another embodiment, please refer to Figure 5 A multi-source data fusion adaptive segmented electricity market reporting system, the system comprising:

[0130] The data acquisition module is used to obtain the medium- and long-term contract electricity volume, forecasted electricity consumption, medium- and long-term average holding price, and forecasted spot price from the database for each time period of the target day.

[0131] The deviation and urgency acquisition module is connected to the data acquisition module and is used to calculate the power deviation and multi-dimensional urgency score for each time period based on the data provided by the data acquisition module.

[0132] An adaptive segmentation module, connected to the deviation and urgency acquisition module, is used to identify deviation patterns based on the sign distribution of the power deviation, and dynamically generate a set of cumulative adjustment amounts containing multiple reporting segments based on the deviation patterns and the multidimensional urgency scores.

[0133] The performance evaluation module, connected to the adaptive segmentation module, is used to simulate each application segment and calculate the average residual absolute deviation, deviation improvement score, and overall score for each application segment.

[0134] The intelligent pricing module, connected to the performance evaluation module, normalizes the comprehensive score and uses the average of the predicted electricity price sequence as the benchmark price. It then uses different pricing formulas to generate initial prices for different declaration segments and performs monotonicity correction to obtain the final declaration price for each declaration segment.

[0135] The output module, connected to the intelligent pricing module, is used to output the declaration scheme, which includes the cumulative adjustment amount, semantic tags, final declaration price, and performance analysis data for each declaration segment.

[0136] In a specific embodiment, the data acquisition module performs step S1, the deviation and urgency acquisition module performs step S2, the adaptive segmentation module performs step S3, the performance evaluation module performs step S4, the intelligent pricing module performs step S5, and the output module performs step S6.

[0137] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0138] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-source data fusion adaptive segmented electricity market declaration method, characterized in that, Includes the following steps: Step S1: Multi-source data acquisition and fusion, acquiring the medium- and long-term contract breakdown electricity volume for each time period of the target day as the basic electricity volume sequence, the load forecast value as the predicted load sequence, the spot market clearing price forecast value as the predicted electricity price sequence, as well as the regulation capacity parameters and the target regulation period set; Step S2: Time-based deviation and urgency analysis. For each time period in the target adjustment time period set, calculate the power consumption deviation, and calculate a multi-dimensional urgency score based on the power consumption deviation, the basic power consumption sequence, the predicted electricity price sequence, whether the time period belongs to the peak period, and whether there is an adjustment demand. Step S3: Adaptive segmented reporting quantity design. Based on the sign distribution of the power deviation of all target adjustment periods, the deviation pattern is identified, and the upper and lower limits of the reporting quantity are dynamically determined by enumerating candidate boundaries and calculating the urgency-weighted coverage. Then, a set of cumulative adjustment quantities containing multiple reporting segments is generated within the upper and lower limits. Step S4: Performance evaluation of the application segment. Simulate the application of each application segment, calculate the average remaining absolute deviation of each application segment for all target adjustment periods, calculate the deviation improvement score based on the average remaining absolute deviation, and calculate the average urgency score for all target adjustment periods. Weight and fuse the deviation improvement score and the average urgency score to obtain the comprehensive score of each application segment. Step S5: Value-based intelligent pricing. After normalizing the comprehensive score, the average of the predicted electricity price series is used as the benchmark price. Different pricing formulas are used to generate initial prices for different declaration segments. When the average urgency score of all target adjustment periods exceeds the preset threshold, the prices of the upward and downward adjustment segments are multiplied by different adjustment coefficients respectively. Finally, monotonicity correction is performed to obtain the final declaration price of each declaration segment. Step S6: Generate the final application plan and output the application plan containing the cumulative adjustment amount, semantic tags, final application price and performance analysis data of each application segment. The performance analysis data includes at least one of the following: average residual deviation, improvement score and comprehensive score.

2. The method according to claim 1, characterized in that, In step S1, the basic electricity consumption sequence is obtained by querying the contract electricity consumption table in the database, summarizing them according to contract type, and then aggregating the time series data into hourly electricity consumption and weighting the average. The predicted electricity price sequence is obtained by querying the electricity price prediction table in the database, averaging it by hour, and using robust statistical methods to remove outliers. The predicted load sequence is obtained by querying the predicted electricity consumption meter in the database, aggregating it by hour, and removing outliers. Regulation capability parameters can be input by the user or obtained from historical regulation capability statistics; The target adjustment period set is a set of several hours that need to be reported and adjusted, as specified by the user or determined by the policy.

3. The method according to claim 1, characterized in that, In step S2, the formula for calculating the multidimensional urgency score Urgency(t) is as follows: Urgency(t) = min(deviation ratio factor + absolute deviation factor + electricity price level factor + peak period factor + adjustment demand factor, 100); Wherein, the deviation scaling factor = min(|Deviation(t)| / |Q_base(t)|×100,100); Absolute deviation factor = min(|Deviation(t)| / 5×50,50); Electricity price level factor = min(P_spot(t) / 20,30); The peak period factor is set to 20 when the period falls within the system load peak, and 0 otherwise. The adjustment demand factor is set to 10 when Deviation(t) ≠ 0, and 0 otherwise. Deviation(t) represents the charge deviation, and the formula for calculating the charge deviation is: Deviation(t)=Q_pred(t)-Q_base(t); Q_pred(t) represents the user's expected electricity consumption, Q_base(t) represents the base electricity consumption, and P_spot(t) represents the predicted electricity price.

4. The method according to claim 1, characterized in that, Step S3 specifically includes: Step S301: Deviation mode identification. If all power deviations are ≥0, it is a full increase mode; if all power deviations are ≤0, it is a full decrease mode; otherwise, it is a mixed mode. Step S302: Calculate net adjustment demand, including total upward adjustment demand and total downward adjustment demand; Step S303: Dynamic boundary generation. Based on the identified pattern, determine the lower limit L and upper limit U of the declared amount. For the full upward or full downward adjustment pattern, the boundary is determined based on the minimum deviation, maximum deviation and adjustment capability parameters. For the mixed pattern, by enumerating multiple candidate boundary combinations, the urgency weighted coverage under each combination is calculated, and the boundary combination that maximizes the coverage is selected as the final (L, U). Step S304: Segment quantity generation and labeling. Within the defined boundary (L, U), five monotonically increasing declaration quantities are generated through linear interpolation, and semantic labels are assigned to each segment based on whether the boundary crosses the zero point.

5. The method according to claim 4, characterized in that, In step S303, the formula for calculating the urgency-weighted coverage is as follows: coverage=Σ_t[w(t)×coverage_rate(t)] / Σ_tw(t); Where w(t) = Urgency(t) / 100, and coverage_rate(t) is the proportion of the deviation in time period t that is covered by the declaration volume of this group; If Deviation(t) > 0 and A[i] ≥ Deviation(t), then the coverage rate is 1; otherwise, take max(A) / Deviation(t). If Deviation(t) < 0 and A[i] ≤ Deviation(t), then the coverage rate is 1; otherwise, take min(A) / Deviation(t). If Deviation(t) = 0: the coverage rate is 1; Where A[i] is the candidate application segment, and min(A) is the maximum candidate application segment.

6. The method according to claim 1, characterized in that, Step S4 specifically includes: Step S401: Simulation application: For each declared segment and each adjustment period, calculate the actual adjustment amount and remaining deviation under the adjustment capacity limit according to the direction and size of the declared segment; Step S402: Calculate the mean residual absolute deviation, mean residual absolute deviation = (1 / |T|)Σ_t |remaining(t)|; Where |T| is the total number of target adjustment periods, and |remaining(t)| is the absolute value of the remaining deviation that still exists in the t-th period after adjustment of a certain reporting segment; Step S403: Calculate the deviation improvement score, deviation improvement score = max(0, 100 - avg_abs_remaining × 10); Where avg_abs_remaining is the average residual absolute deviation; Step S404: Calculate the overall score, Overall Score = Deviation Improvement Score × (1 + Urgency Weight × α); Wherein, the urgency weight = average urgency / 100, average urgency = (1 / |T|)Σ_tUrgency(t), and α is an adjustable coefficient between 0.3 and 0.

7.

7. The method according to claim 1, characterized in that, Step S5 specifically includes: Step 501: Normalization process, normalize the comprehensive scores of all application segments to the [0,1] interval; Step 502: Benchmark price anchoring, using the average of the predicted electricity prices for all target adjustment periods as the market average price; Step 503: Value-driven pricing. The temporary price for the upward adjustment phase = (market average price × first benchmark coefficient) × (first basic competitiveness factor + first adjustment coefficient × (1 - normalized score)). The temporary price for the downward adjustment phase = (market average price × second benchmark coefficient) × (second basic competitiveness factor - second adjustment coefficient × normalized score). Step 504: Monotonicity assurance and minimum interval correction. Force correction is applied to the generated price series to ensure that the bid price strictly decreases as the bid volume increases, and the price difference between adjacent segments is not less than 1% of the highest bid price.

8. A multi-source data fusion adaptive segmented electricity market declaration system, characterized in that, The system includes: The data acquisition module is used to obtain the medium- and long-term contract electricity volume, forecasted electricity consumption, medium- and long-term average holding price, and forecasted spot price from the database for each time period of the target day. The deviation and urgency acquisition module is connected to the data acquisition module and is used to calculate the power deviation and multi-dimensional urgency score for each time period based on the data provided by the data acquisition module. An adaptive segmentation module, connected to the deviation and urgency acquisition module, is used to identify deviation patterns based on the sign distribution of the power deviation, and dynamically generate a set of cumulative adjustment amounts containing multiple reporting segments based on the deviation patterns and the multidimensional urgency scores. The performance evaluation module, connected to the adaptive segmentation module, is used to simulate each application segment and calculate the average residual absolute deviation, deviation improvement score, and overall score for each application segment. The intelligent pricing module, connected to the performance evaluation module, normalizes the comprehensive score and uses the average of the predicted electricity price sequence as the benchmark price. It then uses different pricing formulas to generate initial prices for different declaration segments and performs monotonicity correction to obtain the final declaration price for each declaration segment. The output module, connected to the intelligent pricing module, is used to output the declaration scheme, which includes the cumulative adjustment amount, semantic tags, final declaration price, and performance analysis data for each declaration segment.

Images