Electric power spot market price fluctuation risk early warning method and system, medium and processor

Abstract

The invention relates to the technical field of price fluctuation risk monitoring, and discloses an electric power spot market price fluctuation risk early warning method and system, a medium and a processor, and the method comprises the steps: obtaining electric power load data and electricity price data of an electric power spot market in an electric power system, and carrying out the preprocessing of the data; performing load prediction on the processed power load data through an improved ABC-SA algorithm, and outputting a power load prediction result; inputting the power load prediction result and the electricity price data as sample data into a random forest model for training to obtain a trained electricity price prediction model; inputting real-time power load prediction and electricity price data into the trained electricity price prediction model to obtain an electricity price prediction result; and generating a risk early warning report according to the electricity price prediction result. According to the invention, the accuracy of load prediction can be improved based on the improved ABC-SA prediction method, the monitoring of the price fluctuation of the electricity market is more accurate, and the price fluctuation risk can be dealt with.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity price fluctuation risk monitoring, and in particular to an electricity spot market price fluctuation risk early warning method, system, medium and processor. Background Art

[0002] In the electricity spot market, the price of electricity is determined by the real-time balance of electricity supply and demand. Due to the fact that electricity cannot be stored on a large scale, a small change in the supply and demand of electricity in a short period of time may lead to a large fluctuation in prices.

[0003] Existing research on power load forecasting and price fluctuation risk warning has made significant progress using methods such as time series analysis, machine learning, and intelligent algorithms. However, these methods still face limitations when dealing with large-scale data and complex nonlinear problems in the power market. Specifically, current methods often have difficulty falling into local optimality when dealing with high-dimensional, heterogeneous data, and both efficiency and accuracy are reduced. Summary of the invention

[0004] In view of the problem that the existing power load forecasting and price fluctuation risk warning methods in the prior art have low prediction efficiency and accuracy when processing large-scale data and complex nonlinear problems in the power market, the present invention provides a power spot market price fluctuation risk warning method, system, medium and processor, which can improve the accuracy of load forecasting based on the improved ABC-SA forecasting method, thereby more accurately monitoring power market price fluctuations and more helpful in dealing with price fluctuation risks. The specific technical solution is as follows:

[0005] A method for early warning of price fluctuation risk in a power spot market, comprising:

[0006] Obtaining power load data and electricity price data from the power spot market in the power system, and preprocessing the power load data and electricity price data;

[0007] The improved ABC-SA algorithm is used to perform load forecasting on the processed power load data and output the power load forecasting results;

[0008] The power load forecast results and the processed electricity price data are used as sample data to input into the random forest model for training, and a trained electricity price forecast model is obtained;

[0009] Input the real-time power load forecast and electricity price data into the trained electricity price forecast model to obtain the electricity price forecast result;

[0010] Generate a risk warning report based on the electricity price forecast results.

[0011] Preferably, the preprocessing of the power load data includes:

[0012] The power load data is processed by vertical and horizontal processing methods to obtain the power load value of each sampling point under a preset period;

[0013] The power load value of each sampling point under the preset period is analyzed and processed by the least square method to supplement the missing data of power load;

[0014] The supplemented and complete power load data is normalized.

[0015] Preferably, the electricity price data includes historical day-ahead prices and bidding space data.

[0016] Preferably, the load forecasting is performed on the processed power load data by using the improved ABC-SA algorithm, and the output power load forecasting result includes:

[0017] Step S21, extracting load impact characteristics of the processed power load data by using a rough set algorithm;

[0018] Step S22, setting initialization parameters of the ABC-SA algorithm according to the data scale of the power load forecast;

[0019] Step S23: Combining the load impact characteristics to generate an initial nectar source, and calculating the fitness value of the initial nectar source;

[0020] Step S24: Obtain new nectar sources through neighborhood search by employed bees under preset rules, and calculate the fitness value of the new nectar sources;

[0021] Step S25: according to the fitness value of the initial nectar source and the fitness value of the new nectar source, the nectar source is updated through a preset selection strategy, and the corresponding fitness value is calculated;

[0022] Step S26, determine whether scout bees are generated after the nectar source is updated, if so, find a new nectar source according to a preset strategy, otherwise, output the nectar source with the highest fitness value that indicates the smallest load prediction error;

[0023] Step S27, determine whether the preset iteration termination condition is met after the output nectar source is output. If so, the output nectar source is used as the optimal nectar source to obtain the load prediction result. Otherwise, repeat steps S24 to S26 until the preset iteration termination condition is met.

[0024] Preferably, the new nectar sources obtained by the employed bees through neighborhood search under preset rules include:

[0025] According to the search rules of the ABC algorithm at the beginning of the search, the expression of employed bees searching for new nectar sources is as follows:

[0026]

[0027] In the formula, v ij is the j-th dimension component of the new nectar source; x ij is the j-th dimension component of the current nectar source; x kj represents the j-th dimension component of another nectar source randomly selected from the population; j∈{1,2,…,D}, D is the parameter dimension of the problem to be optimized; i∈{1,2,…,SN}, k∈{1,2,…,SN}, SN is the preset initialization population size; is a random number between [-1,1].

[0028] Preferably, the step of inputting the power load forecasting result and the processed electricity price data as sample data into a random forest model for training to obtain a trained electricity price forecasting model comprises:

[0029] The load forecast results, historical day-ahead prices and bidding space data are used as feature samples and input into the random forest model for training to obtain a trained electricity price prediction model.

[0030] Preferably, generating a risk warning report according to the electricity price forecast result includes:

[0031] Calculate risk assessment indicators based on electricity price forecast results;

[0032] By comparing the risk assessment indicators with the preset risk level threshold, the corresponding risk warning level is obtained.

[0033] A power spot market price fluctuation risk early warning system, applied to the aforementioned power spot market price fluctuation risk early warning method, comprises:

[0034] A data acquisition and processing unit, used to acquire power load data and electricity price data in the power spot market of the power system, and pre-process the power load data and electricity price data;

[0035] A load forecasting unit, used to perform load forecasting on the processed power load data by using an improved ABC-SA algorithm, and output a power load forecasting result;

[0036] A model training unit is used to input the power load forecasting results and the processed electricity price data as sample data into the random forest model for training to obtain a trained electricity price forecasting model;

[0037] The electricity price prediction unit is used to input the real-time power load prediction and electricity price data into the trained electricity price prediction model to obtain the electricity price prediction result;

[0038] The risk warning generation unit is used to generate a risk warning report based on the electricity price forecast results.

[0039] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to implement the aforementioned method for early warning of price fluctuation risks in the spot electricity market.

[0040] A processor is used to run a program, wherein the program executes the aforementioned method for early warning of price fluctuation risks in the spot electricity market when running.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention provides a method for early warning of price fluctuation risks in the electricity spot market by obtaining the electricity load data and electricity price data of the electricity spot market in the power system, and preprocessing the electricity load data and electricity price data; performing load forecasting on the processed electricity load data through the improved ABC-SA algorithm, and outputting the electricity load forecasting result; inputting the electricity load forecasting result and the processed electricity price data as sample data into the random forest model for training, and obtaining a trained electricity price forecasting model; inputting the real-time electricity load forecasting and electricity price data into the trained electricity price forecasting model, and obtaining the electricity price forecasting result; and generating a risk early warning report based on the electricity price forecasting result. The present invention can improve the accuracy of load forecasting based on the improved ABC-SA forecasting method, thereby more accurately monitoring the price fluctuations in the electricity market, and more helpful in coping with price fluctuation risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0044] Figure 1 The present invention is a flow chart of a method for early warning of price fluctuation risks in the electricity spot market.

[0045] Figure 2 The present invention is a flow chart of an embodiment of a method for early warning of price fluctuation risks in the electricity spot market.

[0046] Figure 3 This is a schematic diagram of a price fluctuation risk early warning system for the electricity spot market of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0049] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0050] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0051] See the following examples Figures 1 to 3 .

[0052] The present application provides a method for early warning of price fluctuation risk in a power spot market, including:

[0053] Step S1, obtaining power load data and electricity price data in the power spot market of the power system, and preprocessing the power load data and electricity price data;

[0054] The preprocessing of the power load data comprises:

[0055] The power load data is processed by vertical and horizontal processing methods to obtain the power load value of each sampling point under a preset period;

[0056] Since the collected load data has obvious regularity, that is, the load values ​​at the same time every day are basically the same or similar. If the difference exceeds the design value, the vertical processing method is needed to process the data. Its expression is:

[0057]

[0058] Where r(t) is the fluctuation value at sampling point t; m(t) is the predicted average value of the same sampling point in recent days; t is the sampling point; d is the number of days; Y(d, t) represents the power load value at sampling point t on the dth day.

[0059] Under normal circumstances, the power load is a continuous and stable sequence. When an unstable situation occurs, the data needs to be processed horizontally, as shown in the following formula:

[0060]

[0061] The power load value of each sampling point under the preset period is analyzed and processed by the least square method to supplement the missing data of power load;

[0062] The probability of missing data in power grid data is extremely high, and may even lead to continuous data loss. However, vertical and horizontal processing methods cannot handle consecutive missing values, so it is necessary to fit the same type of data in the previous and next few days. Using the principle of least squares, the missing data is fitted with the normal data of the same type in the days before and after the missing data, as shown in the following formula:

[0063]

[0064] Where m is the upper limit of the summation; k is the index variable of the summation; Δ min is the sum of the squares of the differences; Represents the input data;. The data units in power load forecasting vary significantly, so normalization is used to process the data.

[0065] The supplemented and complete power load data is normalized.

[0066] Normalization can convert data of different dimensions and ranges to the same scale, thereby improving the convergence speed of the model.

[0067] Among them, electricity price data includes historical day-ahead prices and bidding space data. The electricity price data is cleaned and normalized.

[0068] Preprocessed data can improve the operating efficiency and prediction accuracy of subsequent algorithms and avoid the interference of outliers on model training and prediction.

[0069] Step S2: performing load forecasting on the processed power load data by using the improved ABC-SA algorithm, and outputting the power load forecasting result;

[0070] The improved ABC-SA algorithm is used to predict power load data. This algorithm combines the search capability of the ABC algorithm and the global optimization capability of the SA algorithm. It can predict power load more accurately and provide important input data for subsequent electricity price prediction.

[0071] Step S3: input the power load forecast result and the processed electricity price data as sample data into the random forest model for training to obtain a trained electricity price forecast model;

[0072] The obtained power load forecast results and the historical day-ahead prices and bidding space data of the preprocessed electricity price data are organized into a sample data set, in which the power load forecast results are used as feature variables and the electricity price data are used as target variables.

[0073] The random forest model is trained by organizing historical day-ahead prices and bidding space data into sample data sets to obtain a price prediction model based on supply and demand relationships. In the feature selection process, correlation analysis and recursive feature elimination (RFE) are combined to calculate the Pearson correlation coefficient of all possible features related to the target variable, the spot price of electricity. After that, features with correlation coefficients above a certain value are retained for further analysis, and RFE and random forest estimators are applied to determine the most influential features. The RFE process works through a loop in which it iteratively trains and prunes the model, sequentially removing the least important features until an appropriate number of features are selected. Based on cross-validation, a model with a good trade-off between performance and complexity is established.

[0074] The random forest model has high prediction accuracy and anti-overfitting ability. By training the model, the complex nonlinear relationship between power load, historical day-ahead price, bidding space data and electricity price can be learned, thus establishing an accurate electricity price prediction model.

[0075] Step S4: input the real-time power load forecast and electricity price data into the trained electricity price forecast model to obtain the electricity price forecast result;

[0076] Obtaining real-time power load forecast data (which can be continuously updated short-term or ultra-short-term forecast data) and current electricity price data, and inputting them into the trained electricity price forecast model, can make electricity price forecasts based on the current power system status.

[0077] By predicting electricity prices in a timely and accurate manner based on the current operation of the power system, a key basis is provided for subsequent risk warnings.

[0078] Step S5: Generate a risk warning report based on the electricity price forecast result, specifically including:

[0079] Calculate risk assessment indicators based on electricity price forecast results;

[0080] For example, choose volatility indicators, deviation indicators and risk value indicators;

[0081] The volatility index calculates the volatility of the electricity price forecast over a certain period of time. For example, the standard deviation can be used to measure the dispersion of the daily electricity price forecast relative to its mean.

[0082] The deviation index is calculated by calculating the deviation between the electricity price forecast and the historical average electricity price or the expected electricity price (such as the electricity price agreed in the long-term contract).

[0083] The risk value indicates the maximum loss value that may occur in the electricity price within a specific time period in the future at a certain confidence level (such as 95%).

[0084] The selected risk assessment indicators reflect the uncertainty and potential risks of electricity prices from different perspectives. The volatility index can reflect the severity of electricity price fluctuations; the deviation index can reflect the difference between the current electricity price forecast and the historical or expected level, indicating possible abnormal situations; the risk value index gives the maximum possible loss of electricity prices under a certain probability, which is crucial for the assessment of risk tolerance.

[0085] By comparing the risk assessment indicators with the preset risk level threshold, the corresponding risk warning level is obtained.

[0086] The calculated risk assessment indicators are compared with the corresponding thresholds, and the final risk warning level is determined by comprehensively considering various indicators. For example, if the volatility indicator shows medium risk, the deviation indicator shows low risk, and the risk value indicator shows high risk, it is necessary to comprehensively weigh or determine the overall risk warning level as high risk based on more important indicators (such as risk value for enterprises that are sensitive to risk tolerance).

[0087] A method for early warning of price fluctuation risks in the electricity spot market in an embodiment of the present application uses an improved ABC-SA algorithm for load forecasting, combining the search capability of the artificial bee colony algorithm (ABC) and the global optimization capability of the simulated annealing algorithm (SA) to accurately predict the power load. The power load forecast results and the processed electricity price data are input as sample data into the random forest model for training, and the obtained electricity price forecast model can more accurately predict future electricity price trends. A risk warning report is generated based on the electricity price forecast results. By calculating risk assessment indicators (such as electricity price volatility, deviation, risk value, etc.) and comparing them with preset risk level thresholds, the risk level of electricity price fluctuations can be determined. The present application can improve the accuracy of load forecasting based on the improved ABC-SA forecasting method, thereby more accurately monitoring price fluctuations in the electricity market and more helpful in dealing with price fluctuation risks.

[0088] Specifically, in a preferred implementation of the present application, the load forecasting is performed on the processed power load data by using the improved ABC-SA algorithm, and the output power load forecasting result includes:

[0089] Step S21, extracting load impact characteristics of the processed power load data by using a rough set algorithm;

[0090] The final load impact features can be extracted based on the result of attribute reduction. These features will be used as inputs for subsequent load forecasting models. For example, the average load within a certain period of time, the daily change rate of load, and weather-related features may be extracted.

[0091] Step S22, setting initialization parameters of the ABC-SA algorithm according to the data scale of the power load forecast;

[0092] Determine population size:

[0093] The size of the population is determined by analyzing the amount of data for power load forecasting. For example, the population size is set at 100.

[0094] Set the honey source to limit the number of mining times:

[0095] Based on the complexity of the data and the expected difficulty of searching, determine the maximum number of times the honey source is allowed to be mined without improvement, for example 30 times.

[0096] Determine the maximum number of iterations:

[0097] For large-scale data, more iterations may be required to ensure that the optimal solution is found. For example, the maximum number of iterations can be set to 200 and a suitable value can be found through trial and error.

[0098] Initialize the relevant parameters of simulated annealing:

[0099] The initial temperature of simulated annealing can be set. For example, for complex power load forecasting problems, the initial temperature can be set to 50.

[0100] Determine the temperature drop rate and termination temperature. The temperature drop rate determines the speed of algorithm convergence, and the termination temperature indicates the condition for the algorithm to stop. For example, the temperature drop rate can be set to 0.8 times the temperature of the previous round in each iteration, and the termination temperature is set to 0.5.

[0101] Step S23: Combining the load impact characteristics to generate an initial nectar source, and calculating the fitness value of the initial nectar source;

[0102] The extracted load impact features are encoded and represented as vectors. For example, if there are three load impact features (such as daily average load, temperature, and date type).

[0103] For each initial nectar source, the fitness value is calculated according to its characteristic value. For example, the mean square error (MSE) can be used as the fitness function to calculate the mean square error between the predicted load and the actual load. The smaller the error, the higher the fitness value.

[0104] Step S24: Obtain new nectar sources through neighborhood search by employed bees under preset rules, and calculate the fitness value of the new nectar sources;

[0105] According to the search rules of the ABC algorithm at the beginning of the search, the expression of employed bees searching for new nectar sources is as follows:

[0106]

[0107] In the formula, v ij is the j-th dimension component of the new nectar source; x ij is the j-th dimension component of the current nectar source; x kj represents the j-th dimension component of another nectar source randomly selected from the population; j∈{1,2,…,D}, D is the parameter dimension of the problem to be optimized; i∈{1,2,…,SN}, k∈{1,2,…,SN}, SN is the preset initialization population size; is a random number between [-1,1].

[0108] According to the above rules, hired bees search for the nectar sources they are responsible for and generate new nectar sources. Each hired bee performs this operation independently, and the number of new nectar sources generated is the same as the number of hired bees.

[0109] For the newly generated nectar source, its fitness value is calculated in the same way as the previous fitness function (such as mean square error).

[0110] Step S25: according to the fitness value of the initial nectar source and the fitness value of the new nectar source, the nectar source is updated through a preset selection strategy, and the corresponding fitness value is calculated;

[0111] If the fitness value of the new nectar source is better than the fitness value of the initial nectar source, the initial nectar source will be replaced by the new nectar source unconditionally; if the fitness value of the new nectar source is worse than the fitness value of the initial nectar source, whether to accept the new nectar source will be decided based on the probability acceptance criterion of the simulated annealing algorithm.

[0112] According to the above selection strategy, the nectar source is updated. If the new nectar source is accepted, the initial nectar source is replaced with the new nectar source; if not accepted, the initial nectar source is retained. For the updated nectar source (whether replaced or retained), its fitness value is recalculated to reflect the latest status of the nectar source.

[0113] Step S26, determine whether scout bees are generated after the nectar source is updated, if so, find a new nectar source according to a preset strategy, otherwise, output the nectar source with the highest fitness value that indicates the smallest load prediction error;

[0114] Check the number of times each nectar source has been mined (searched and updated) and compare it with the preset nectar source limit. If the number of times a nectar source has been mined exceeds the limit, it is considered that the nectar source has fallen into a local optimum and needs to send scout bees to find a new nectar source.

[0115] When scout bees are needed, they search for new nectar sources according to preset strategies, for example, randomly generating new nectar sources in the entire search space, or using other heuristic methods to search for nectar sources that may jump out of the local optimum in a larger range.

[0116] If no nectar source has reached the mining limit, that is, no scout bees are needed, then among all current nectar sources, the nectar source with the highest fitness value (the smallest load prediction error) is found and prepared for output.

[0117] Step S27, determine whether the preset iteration termination condition is met after the output nectar source is output. If so, the output nectar source is used as the optimal nectar source to obtain the load prediction result. Otherwise, repeat steps S24 to S26 until the preset iteration termination condition is met.

[0118] Check whether the preset iteration termination conditions are met. The termination conditions generally include reaching the maximum number of iterations or the simulated annealing temperature dropping to the termination temperature; if the termination conditions are met, the currently output nectar source is used as the optimal nectar source. According to the load impact characteristics contained in the optimal nectar source, the power load forecast result is calculated and output; if the termination conditions are not met, return to step S24 and continue the neighborhood search of employed bees, nectar source update, scout bee judgment and other operations until the termination conditions are met.

[0119] The embodiment of the present application further provides a power spot market price fluctuation risk early warning system, which is applied to the aforementioned power spot market price fluctuation risk early warning method, comprising:

[0120] A data acquisition and processing unit, used to acquire power load data and electricity price data in the power spot market of the power system, and pre-process the power load data and electricity price data;

[0121] A load forecasting unit, used to perform load forecasting on the processed power load data by using an improved ABC-SA algorithm, and output a power load forecasting result;

[0122] A model training unit is used to input the power load forecasting results and the processed electricity price data as sample data into the random forest model for training to obtain a trained electricity price forecasting model;

[0123] The electricity price prediction unit is used to input the real-time power load prediction and electricity price data into the trained electricity price prediction model to obtain the electricity price prediction result;

[0124] The risk warning generation unit is used to generate a risk warning report based on the electricity price forecast results.

[0125] The functional explanation of each unit in a power spot market price fluctuation risk warning system of this embodiment corresponds to the explanation in an power spot market price fluctuation risk warning method embodiment, and the technical effects correspond, so they will not be repeated.

[0126] An embodiment of the present application also provides a computer-readable storage medium, which includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to implement the aforementioned method for early warning of price fluctuation risks in the spot electricity market.

[0127] The technical effect of this embodiment is the same as the technical effect of a method for early warning of price fluctuation risks in the electricity spot market in an embodiment, and will not be repeated here.

[0128] The present invention can be used in many general or special computer system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like.

[0129] An embodiment of the present application also provides a processor, which is used to run a program, wherein the program executes the aforementioned method for early warning of price fluctuation risks in the spot electricity market when running.

[0130] The technical effect of this embodiment is the same as the technical effect of the method for early warning of price fluctuation risks in the electricity spot market in embodiment 1, and will not be repeated here.

[0131] The processor in this embodiment may be a central processing unit (CPU), a controller, a microcontroller, or other data processing chips.

[0132] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0133] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0134] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0135] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for early warning of price fluctuation risk in the electricity spot market, characterized in that: include: Obtaining power load data and electricity price data from the power spot market in the power system, and preprocessing the power load data and electricity price data; The improved ABC-SA algorithm is used to perform load forecasting on the processed power load data and output the power load forecasting results; The power load forecast results and the processed electricity price data are used as sample data to input into the random forest model for training, and a trained electricity price forecast model is obtained; Input the real-time power load forecast and electricity price data into the trained electricity price forecast model to obtain the electricity price forecast result; Generate a risk warning report based on the electricity price forecast results.

2. The method for early warning of price fluctuation risk in the electricity spot market according to claim 1, characterized in that: The preprocessing of the power load data comprises: The power load data is processed by vertical and horizontal processing methods to obtain the power load value of each sampling point under the preset period; the power load value of each sampling point under the preset period is analyzed and processed by the least squares method to supplement the missing power load data; the supplemented power load data is normalized.

3. The method for early warning of price fluctuation risk in the electricity spot market according to claim 2, characterized in that: The electricity price data includes historical day-ahead prices and bidding space data.

4. A method for early warning of price fluctuation risk in the electricity spot market according to claim 3, characterized in that: The improved ABC-SA algorithm is used to perform load forecasting on the processed power load data, and the output power load forecasting results include: Step S21, extracting load impact characteristics of the processed power load data by using a rough set algorithm; Step S22, setting initialization parameters of the ABC-SA algorithm according to the data scale of the power load forecast; Step S23: Combining the load impact characteristics to generate an initial nectar source, and calculating the fitness value of the initial nectar source; Step S24: Obtain new nectar sources through neighborhood search by employed bees under preset rules, and calculate the fitness value of the new nectar sources; Step S25: according to the fitness value of the initial nectar source and the fitness value of the new nectar source, the nectar source is updated through a preset selection strategy, and the corresponding fitness value is calculated; Step S26, determine whether scout bees are generated after the nectar source is updated, if so, find a new nectar source according to a preset strategy, otherwise, output the nectar source with the highest fitness value that indicates the smallest load prediction error; Step S27, determine whether the preset iteration termination condition is met after the output nectar source is output. If so, the output nectar source is used as the optimal nectar source to obtain the load prediction result. Otherwise, repeat steps S24 to S26 until the preset iteration termination condition is met.

5. A method for early warning of price fluctuation risk in the electricity spot market according to claim 4, characterized in that: The new nectar sources obtained by the employed bees through neighborhood search under preset rules include: According to the search rules of the ABC algorithm at the beginning of the search, the expression of employed bees searching for new nectar sources is as follows: In the formula, v ij is the j-th dimension component of the new nectar source; x ij is the j-th dimension component of the current nectar source; x kj represents the j-th dimension component of another nectar source randomly selected from the population; j∈{1,2,…,D}, D is the parameter dimension of the problem to be optimized; i∈{1,2,…,SN}, k∈{1,2,…,SN}, SN is the preset initialization population size; is a random number between [-1,1].

6. The method for early warning of price fluctuation risk in the electricity spot market according to claim 1, characterized in that: The power load forecasting results and the processed electricity price data are input as sample data into the random forest model for training to obtain a trained electricity price forecasting model, which includes: The load forecast results, historical day-ahead prices and bidding space data are used as feature samples and input into the random forest model for training to obtain a trained electricity price prediction model.

7. A method for early warning of price fluctuation risk in the electricity spot market according to claim 6, characterized in that: The generating of risk warning report according to the electricity price forecast result includes: Calculate risk assessment indicators based on electricity price forecast results; By comparing the risk assessment indicators with the preset risk level threshold, the corresponding risk warning level is obtained.

8. An early warning system for price fluctuation risk in the electricity spot market, characterized in that: A method for early warning of price fluctuation risk in the electricity spot market as applied to any one of claims 1 to 7, comprising: A data acquisition and processing unit, used to acquire power load data and electricity price data in the power spot market of the power system, and pre-process the power load data and electricity price data; A load forecasting unit, used to perform load forecasting on the processed power load data by using an improved ABC-SA algorithm, and output a power load forecasting result; A model training unit is used to input the power load forecasting results and the processed electricity price data as sample data into the random forest model for training to obtain a trained electricity price forecasting model; The electricity price prediction unit is used to input the real-time power load prediction and electricity price data into the trained electricity price prediction model to obtain the electricity price prediction result; The risk warning generation unit is used to generate a risk warning report based on the electricity price forecast results.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein: When the program is running, the device where the computer-readable storage medium is located is controlled to execute a method for early warning of price fluctuation risks in the electricity spot market as described in any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein when the program is run, an electricity spot market price fluctuation risk early warning method as described in any one of claims 1 to 7 is executed.

Images