Electric power bidding prediction model optimization method and device and electric power bidding prediction method and device

Abstract

The invention provides an electric power bidding prediction model optimization and electric power bidding prediction method and device, and relates to the technical field of electric power market and electric power system optimization, and the method comprises the steps: obtaining the unit configuration data of a target object and the historical load information of a user; analyzing the user historical load information and the unit configuration data to determine a target factor; establishing an initial electric power bidding prediction model of the target object based on the target factor and a preset balance constraint condition; and performing iterative adjustment on the model parameters of the initial electric power bidding prediction model based on a firefly algorithm until a preset end condition is met, and generating a target electric power bidding prediction model. Therefore, through the method of the scheme, the accuracy of electric power bidding prediction can be improved, the accuracy can be used as a basis for optimizing resource allocation, and meanwhile, the competitiveness of users in the electric power market can be enhanced.

Description

Technical Field

[0001] The present disclosure relates to the technical field of power market and power system optimization, and in particular to a power bidding prediction model optimization and a power bidding prediction method and device. Background Art

[0002] Since the launch of the new round of power system reform, various parts of the country have successively formulated corresponding power trading rules to promote the reform process. However, since the contracted electricity volume in medium- and long-term power transactions is based on the planned value, this inevitably leads to systematic errors. Therefore, before the establishment of the power spot market, most regions implemented a power deviation assessment mechanism as a transitional solution. However, many power sales companies regard the assumption of deviation assessment as a marketing strategy to attract customers, which makes the deviation assessment rules a severe challenge to the profitability of power sales companies.

[0003] In the electricity market, as the market environment changes, electricity sellers constantly adjust their bidding strategies to better adapt to the market environment and maximize their own interests. The adjustment of each electricity seller's own strategy will affect the bidding strategies of other electricity sellers, and then affect the changes in the environment of the entire electricity market. Therefore, the electricity market is essentially a complex adaptive system. Agent-based modeling is a typical representative of this type of modeling method. The idea of ​​multi-agent modeling is to regard the participants in the system as an agent and give it intelligence, so that the system participants interact with each other to obtain the corresponding system model. Summary of the invention

[0004] The present disclosure aims to solve one of the technical problems in the related art at least to some extent.

[0005] To this end, the first objective of the present disclosure is to propose a method for optimizing an electricity bidding prediction model.

[0006] The second objective of the present disclosure is to propose a method for predicting electricity bidding.

[0007] The third objective of the present disclosure is to propose an electricity bidding prediction model optimization device.

[0008] The fourth objective of the present disclosure is to provide an electricity bidding prediction device.

[0009] A fifth objective of the present disclosure is to provide an electronic device.

[0010] A sixth object of the present disclosure is to provide a non-transitory computer-readable storage medium.

[0011] A seventh objective of the present disclosure is to provide a computer program product.

[0012] To achieve the above-mentioned purpose, the first aspect of the present disclosure proposes a method for optimizing an electricity bidding prediction model, including: obtaining unit configuration data and user historical load information of a target object; parsing the user historical load information and the unit configuration data to determine a target factor; establishing an initial electricity bidding prediction model of the target object based on the target factor and a preset balance constraint condition; iteratively adjusting the model parameters of the initial electricity bidding prediction model based on the firefly algorithm until a preset end condition is met, thereby generating a target electricity bidding prediction model.

[0013] According to one embodiment of the present disclosure, the parsing of the user's historical load information and the unit configuration data to determine the target factor includes: performing semantic processing on the user's historical load information to generate multiple candidate factors; calculating contribution values ​​of the multiple candidate factors to electricity bidding; and determining the target factor from the multiple candidate factors based on the contribution values.

[0014] According to one embodiment of the present disclosure, the initial electricity bidding prediction model of the target object is established based on the target factor and the preset balance constraint condition, including: establishing the target function of the target object based on the target factor; establishing the initial electricity bidding prediction model of the target object based on the target function and the balance constraint condition.

[0015] According to one embodiment of the present disclosure, the target factors include the actual electricity sales to users, the actual electricity sales price to users, the long-term contract electricity price, the electricity purchase price from the power market, the electricity purchase volume from the power market, and the monthly decomposition value of the annual long-term contract electricity purchased by the power sales company. The target function is: Wherein, t is the bidding time interval, and n t is the actual amount of electricity sold to the user, and R i,t is the actual electricity price sold to users, and M contact,t is the monthly decomposition value of the annual long-term contract electricity purchased by the power sales company. contact,t is the long-term contract electricity price, and C buy,t is the price of electricity purchased from the power market, and M buy,t is the amount of electricity purchased from the electricity market, the maxP total Profit data for the target object.

[0016] According to an embodiment of the present disclosure, the balancing constraint condition includes an electricity balancing constraint condition, an electricity purchase constraint condition and a quotation condition.

[0017] According to one embodiment of the present disclosure, the model parameters of the initial electricity bidding prediction model are iteratively adjusted based on the firefly algorithm until a preset end condition is met to generate a target electricity bidding prediction model, including: determining the model parameter range of the initial electricity bidding prediction model, and randomly generating an initial firefly factor set, wherein the position of any candidate firefly factor in the firefly factor set represents a set of model parameters of the initial electricity bidding prediction model; iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm, and determining whether the updated candidate firefly factors meet the end condition; repeating the above process of iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm until the end condition is met to generate a target electricity bidding prediction model.

[0018] To achieve the above-mentioned purpose, the second aspect embodiment of the present disclosure proposes a method for optimizing an electricity bidding prediction model, comprising: predicting a user's electricity consumption through a user's electricity consumption model to generate a usage prediction value, wherein the user's electricity consumption model is an LSTM model; inputting the usage prediction value into a target electricity bidding prediction model to generate an electricity bidding value of a target object, wherein the target electricity bidding prediction model is generated by the electricity bidding prediction model optimization method as described in the first aspect embodiment.

[0019] To achieve the above-mentioned purpose, the third aspect embodiment of the present disclosure proposes an electricity bidding prediction model optimization device, including: an acquisition module, used to acquire the unit configuration data and user historical load information of the target object; a parsing module, used to parse the user historical load information and the unit configuration data to determine the target factor; an establishment module, used to establish the initial electricity bidding prediction model of the target object based on the target factor and the preset balance constraint condition; a generation module, used to iteratively adjust the model parameters of the initial electricity bidding prediction model based on the firefly algorithm until the preset end condition is met, and generate a target electricity bidding prediction model.

[0020] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present disclosure proposes an electricity bidding prediction device, including: a prediction module, used to predict the user's electricity consumption through the user's electricity consumption model to generate a usage prediction value, wherein the user's electricity consumption model is an LSTM model; an input module, used to input the usage prediction value into a target electricity bidding prediction model to generate an electricity bidding value of a target object, wherein the target electricity bidding prediction model is generated by the electricity bidding prediction model optimization method as described in the first aspect embodiment.

[0021] To achieve the above-mentioned purpose, the fifth aspect embodiment of the present disclosure proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to implement the electricity bidding prediction model optimization method as described in the first aspect embodiment of the present disclosure, or the electricity bidding prediction model optimization method as described in the second aspect embodiment.

[0022] To achieve the above-mentioned purpose, the sixth aspect embodiment of the present disclosure proposes a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to implement the electricity bidding prediction model optimization method as described in the first aspect embodiment of the present disclosure, or the electricity bidding prediction model optimization method as described in the second aspect embodiment.

[0023] To achieve the above-mentioned purpose, the seventh aspect embodiment of the present disclosure proposes a computer program product, including a computer program, which, when executed by a processor, is used to implement the electricity bidding prediction model optimization method as described in the first aspect embodiment of the present disclosure, or the electricity bidding prediction model optimization method as described in the second aspect embodiment.

[0024] Therefore, the method of this scheme can improve the accuracy of electricity bidding prediction, and can be used as a basis for optimizing resource allocation, while enhancing the competitiveness of users in the electricity market. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic diagram of an optimization method of a power bidding prediction model according to an embodiment of the present disclosure;

[0026] Figure 2 is a schematic diagram of another power bidding prediction model optimization method according to an embodiment of the present disclosure;

[0027] Figure 3 is a schematic diagram of another power bidding prediction model optimization method according to an embodiment of the present disclosure;

[0028] Figure 4 is a schematic diagram of another power bidding prediction model optimization method according to an embodiment of the present disclosure;

[0029] Figure 5 is a schematic diagram of a power bidding prediction method according to an embodiment of the present disclosure;

[0030] Figure 6 It is a schematic diagram of an electric power bidding prediction model optimization device according to one embodiment of the present disclosure;

[0031] Figure 7is a schematic diagram of a power bidding prediction device according to an embodiment of the present disclosure;

[0032] Figure 8 is a schematic diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0033] Embodiments of the present disclosure are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0034] The acquisition, storage, use, and processing of data in the technical solution disclosed in this disclosure are in compliance with the relevant provisions of relevant laws and regulations.

[0035] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be considered as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0036] Figure 1 is a schematic diagram of an optimization method for an electricity bidding prediction model according to an embodiment of the present disclosure, such as Figure 1 As shown, the power bidding prediction model optimization method includes the following steps:

[0037] S101, obtaining unit configuration data and user historical load information of the target object.

[0038] The target object is the object for which the power bidding forecast is conducted in this scheme. The target object can be of various types and is not limited here. For example, the target object can be a power sales company, a power plant, etc.

[0039] The electricity bidding prediction model optimization method of the embodiment of the present application can be applied to the scenario of electricity trading. The executor of the electricity bidding prediction model optimization of the embodiment of the present application can be the electricity bidding prediction model optimization device of the embodiment of the present application, and the electricity bidding prediction model optimization device can be set on an electronic device.

[0040] In the disclosed embodiment, the unit configuration data may include a variety of data, which is not limited here. For example, it may include the unit's operating parameters, performance indicators, status monitoring, etc.

[0041] The household historical load information is very important data in the power system, which reflects the user's power consumption pattern in the past period of time. The household historical load information can include a variety of data, which is not limited here, for example, the user's power consumption, power purchase price, power consumption habits, etc.

[0042] There are many methods for obtaining the unit configuration data and user historical load information of the target object, which can be specifically limited according to actual design needs.

[0043] In one possible implementation, the unit configuration data and the user's historical load information may be obtained by acquiring the historical operating data or transaction data of the target object.

[0044] In another possible implementation method, the data can also be obtained by analyzing the user's historical power usage or the operation log of the unit data.

[0045] S102, analyzing the user's historical load information and unit configuration data to determine the target factor.

[0046] It should be noted that the unit configuration data and user historical load information may include a variety of different data, but not all data will have an impact on the profitability of the target object.

[0047] Therefore, before building a model, it is necessary to first determine one or more target factors that have the greatest impact on the profitability of the target object. This can improve the accuracy and prediction effect of the subsequent model.

[0048] In the embodiment of the present disclosure, there are many methods for parsing the user's historical load information and unit configuration data to determine the target factor, and no limitation is made here.

[0049] In one possible implementation, the user historical load information may be calculated by an analytical algorithm to determine a target factor from a plurality of user historical load information and unit configuration data.

[0050] In another possible implementation, the user's historical load information and unit configuration data can also be input into the target factor determination model to determine the target factor. The target factor determination model is a pre-trained model and is stored in the storage space of the electronic device for easy retrieval when needed.

[0051] S103: establishing an initial power bidding prediction model for the target object based on the target factors and preset balance constraints.

[0052] S104, iteratively adjusting the model parameters of the initial power bidding prediction model based on the firefly algorithm until a preset end condition is met, thereby generating a target power bidding prediction model.

[0053] It should be noted that the Firefly Algorithm is an important part of the modeling and simulation methods based on multi-agent systems, and is often used in the modeling of complex adaptive systems. As an optimization algorithm, the Firefly Algorithm regards the optimization process as an evolutionary process, and updates the population based on the principle of mutual attraction between individuals based on the dominance relationship, thereby driving the population to evolve in the direction of the optimal solution until the optimal value is finally found.

[0054] In the disclosed embodiment, the end condition is designed in advance and can be changed according to actual design needs, and no limitation is made here. For example, the end condition can be that the iteration reaches a preset number of times, or the iteration result meets the preset condition, etc.

[0055] In the disclosed embodiment, the unit data setting of the target object and the user's historical load information are first obtained, and then the user's historical load information is parsed to determine the target factor, and then the initial power bidding prediction model of the target object is established based on the target factor and the preset balance constraint condition, and finally the model parameters of the initial power bidding prediction model are iteratively adjusted based on the firefly algorithm until the preset end condition is met to generate the target power bidding prediction model. Therefore, through the method of this scheme, the accuracy of power bidding prediction can be improved, and it can be used as a basis for optimizing resource allocation, and at the same time, the competitiveness of users in the power market can be enhanced.

[0056] In the above embodiment, the user's historical load information and unit configuration data are analyzed to determine the target factor, and the target factor can also be determined by Figure 2 Explaining further, the method includes:

[0057] S201, semantically processing the user's historical load information to generate multiple candidate factors.

[0058] In the embodiment of the present disclosure, semantic processing may be performed on the user's historical load information through a natural language algorithm to generate a plurality of candidate factors.

[0059] The natural language algorithm is designed in advance and can be changed according to actual design needs, and no limitation is made here.

[0060] S202, calculating contribution values ​​of multiple candidate factors to electricity bidding.

[0061] It should be noted that the contribution value is a parameter value for determining the degree of influence of a candidate factor on electricity bidding. In the embodiment of the present disclosure, multiple candidate factors can be calculated using a preset contribution value calculation algorithm to obtain the contribution values ​​of the multiple candidate factors.

[0062] S203: Determine a target factor from multiple candidate factors based on the contribution value.

[0063] In the embodiment of the present disclosure, after the contribution values ​​of the multiple candidate factors are acquired, the contribution values ​​of the multiple candidate factors may be compared with a contribution threshold to determine a target factor from the multiple candidate factors.

[0064] It should be noted that the contribution threshold is a critical value for determining a candidate factor as a target factor. When the contribution value of a candidate factor is greater than the contribution threshold, the candidate factor can be considered as the target factor.

[0065] In the embodiment of the present disclosure, the user's historical load information is first semantically processed to generate multiple candidate factors, then the contribution values ​​of the multiple candidate factors to the power bidding are calculated, and finally the target factor is determined from the multiple candidate factors based on the contribution values. Thus, by calculating the contribution values, the target factor is determined from the multiple candidate factors, which can provide an accurate data basis for the subsequent model establishment.

[0066] In the above embodiment, the initial power bidding prediction model of the target object is established based on the target factor and the preset balance constraint condition, and the Figure 3 Explaining further, the method includes:

[0067] S301, establishing an objective function of a target object based on the target factor.

[0068] It should be noted that the target factors include the actual electricity sales to users, the actual electricity sales price to users, the long-term contract electricity price, the electricity purchase price from the power market, the electricity purchase volume from the power market, and the monthly decomposition value of the long-term contract electricity purchased by the power sales company in the year. The target function is:

[0069]

[0070] Among them, t is the bidding time interval, η t To actually sell electricity to users, R i,t The actual electricity price sold to users, M contact,t is the monthly decomposition value of the annual long-term contract electricity purchased by the power sales company, C contact,t is the long-term contract electricity price, C buy,t is the price of electricity purchased from the electricity market, M buy,t To purchase electricity from the electricity market, maxP total Profit data for the target object.

[0071] It should be noted that the long-term contract price refers to the electricity price agreed upon in a long-term agreement (Long-term Agreement, referred to as long-term agreement). This type of contract is usually used to ensure the stability of electricity supply and consumption between the power supplier and the power user in the future, and also provides a certain degree of price protection for both parties to avoid the risks brought by market price fluctuations.

[0072] The monthly breakdown value of the annual long-term contract electricity volume purchased by the power sales company refers to the total electricity volume agreed in the annual long-term agreement (long-term agreement), which is allocated to specific values ​​for each month according to certain rules.

[0073] S302: Establishing an initial electricity bidding prediction model for a target object based on an objective function and balance constraints.

[0074] In the embodiment of the present disclosure, the balancing constraint conditions include power balancing constraint conditions, power purchase constraint conditions and quotation conditions.

[0075] Power balance constraints:

[0076]

[0077] Among them, M punish,t is the monthly deviation electricity. If the deviation electricity is positive, the assessment electricity price is twice the monthly competitive electricity priceξ t =2C buy,t If the deviation power is negative, the assessment power price is 3 times the monthly competitive power priceξ t =3C buy,t The sum of the purchased electricity and the deviation electricity should be equal to the actual electricity consumption of the user, and the deviation assessment electricity shall be supplied by a special generator set.

[0078] Power purchase constraints:

[0079]

[0080]

[0081]

[0082] in, is the predicted value of power load of user i at time t, is the mean value of load forecast; is the upper bound of the prediction deviation, The purchased electricity should be greater than the predicted electricity to prevent deviation assessment due to insufficient electricity.

[0083] Quote constraints:

[0084]

[0085] Among them, in order to prevent vicious competition, maximum and minimum bid values ​​will be set during the bidding process in the electricity market to guide the healthy development of the market. is the minimum value of the quotation, It is the maximum value of the quotation. It should be noted that the maximum and minimum values ​​are designed in advance and can be changed according to actual design needs. There is no limitation here.

[0086] In the embodiment of the present disclosure, the objective function of the target object is first established based on the target factor, and then the initial power bidding prediction model of the target object is established based on the objective function and the balance constraint condition. Thus, by setting the objective function and the balance constraint condition, the target power bidding prediction model finally generated can be adapted to the current market rules, and at the same time, the practicality and prediction accuracy of the target power bidding prediction model can be improved, thereby improving the user experience.

[0087] In the above embodiment, the model parameters of the initial power bidding prediction model are iteratively adjusted based on the firefly algorithm until the preset end condition is met to generate the target power bidding prediction model. Figure 4 Explaining further, the method includes:

[0088] S401, determining a model parameter range of an initial electricity bidding prediction model, and randomly generating an initial firefly factor set, wherein the position of any candidate firefly factor in the firefly factor set represents a set of model parameters of the initial electricity bidding prediction model.

[0089] In the embodiments of the present disclosure, the model parameter range can be changed according to actual design needs, and no limitation is made here. Different target objects may have different corresponding model parameter ranges.

[0090] S402, iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm, and determining whether the updated candidate firefly factors meet the termination condition.

[0091] It should be noted that before iteratively updating the positions of candidate firefly factors in the firefly factor set based on the firefly algorithm, the parameters of the firefly algorithm need to be initialized.

[0092] For example, the following values ​​can be initialized.

[0093] Number of fireflies ((n)): determines the number of points initially randomly distributed in the search space.

[0094] Maximum number of iterations (max_iter): Sets the maximum number of running cycles of the algorithm.

[0095] Attraction coefficient: the maximum attraction value between fireflies (when the distance is 0).

[0096] Light absorption coefficient: affects how quickly attraction decays with distance.

[0097] Random move step size: controls the degree of random exploration.

[0098] In the embodiment of the present disclosure, the process of iteratively updating the positions of candidate firefly factors in the firefly factor set based on the firefly algorithm can be as follows:

[0099] For each pair of fireflies (i) and (j):

[0100] If the brightness (fitness) of a firefly (j) is better than that of (i), the attraction is calculated based on the distance (r) between them.

[0101] The attraction of firefly j to firefly i is defined as follows:

[0102]

[0103] Among them, β0 is the maximum attraction, which is equal to the attraction at the light source r = 1, usually β0 = 1; γ is the light absorption coefficient, usually γ∈[0.01,100]; r ij is the Euclidean distance from firefly i to firefly j.

[0104] In the current environment, if there is a Pareto dominance of firefly j over firefly i, then:

[0105]

[0106] Among them, t is the current iteration number; x i 、x j are the spatial positions of fireflies i and j respectively; β ij (r ij ) is the attraction between fireflies i and j; α t is a constant, generally α∈[0,1]; is a vector of random numbers obtained from a Gaussian, uniform, or other distribution.

[0107] If there is no Pareto dominance of firefly j over firefly i, then:

[0108]

[0109] in, It is the optimal value obtained by converting multiple objective functions into a single objective function in a randomly weighted sum manner.

[0110] S403, repeating the above process of iteratively updating the positions of candidate firefly factors in the firefly factor set based on the firefly algorithm until the end condition is met, and generating a target power bidding prediction model.

[0111] In the disclosed embodiment, the model parameter range of the initial electricity bidding prediction model is first determined, and an initial firefly factor set is randomly generated, wherein the position of any candidate firefly factor in the firefly factor set represents a set of model parameters of the initial electricity bidding prediction model, and then the positions of the candidate firefly factors in the firefly factor set are iteratively updated based on the firefly algorithm, and it is determined whether the updated candidate firefly factors meet the end condition, and finally the above process of iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm is repeated until the end condition is met to generate the target electricity bidding prediction model. Iterative optimization of the model parameters of the initial electricity bidding prediction model through the firefly algorithm not only improves the performance and reliability of the model, but also brings significant competitive advantages and economic benefits to power suppliers. This method also provides strong support for the refined management and intelligent development of the power industry.

[0112] Figure 5 is a schematic diagram of a power bidding prediction method according to an embodiment of the present disclosure, such as Figure 5 As shown, the power bidding prediction method includes the following steps:

[0113] S501, predicting the user's power consumption by using the user's power consumption model to generate a usage prediction value, wherein the user's power consumption model is an LSTM model.

[0114] In the disclosed embodiment, the user power consumption model is trained in advance and stored in the storage space of the electronic device to facilitate retrieval and use when needed.

[0115] It should be noted that the full name of the LSTM model is the "Long Short-Term Memory" model. This is a special type of recurrent neural network (RNN), which is mainly used to process and predict long-term dependencies in time series data. LSTM controls the flow of information by introducing a special structure - memory cell and input gate, forget gate, and output gate, thereby effectively solving the gradient vanishing or gradient exploding problem encountered by traditional RNN when processing long sequences.

[0116] The LSTM gate structure includes four key units: the forget gate f t , update gate u t , cell status and output gate o t Each time step has three inputs: external input x at time t t , LSTM unit output h at time t-1t-1 , cell state C at time t-1 t-1 The above input data is processed by the LSTM gating unit to obtain the output h t and cell state C t The gate control unit output calculation process is as follows:

[0117] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0118] u t =σ(W u ·[h t-1 ,x t ]+b u )

[0119]

[0120]

[0121] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0122] h t =o t tanh(C t )

[0123] Where tanh represents the hyperbolic tangent function and σ represents the activation equation. f , W u , W c , W o and b f 、b u 、b c 、b o They represent the weights and bias matrices of the forget gate, update gate, cell state, and output gate, respectively. The initial values ​​of the above weights and bias matrices are randomly set and obtained through data training.

[0124] By predicting users' electricity consumption through the LSTM model, the accuracy of the prediction of users' future electricity consumption can be improved and the impact of uncertainty factors on the prediction results can be reduced.

[0125] S502: Input the usage prediction value into the target power bidding prediction model to generate the power bidding value of the target object.

[0126] It should be noted that the target power bidding prediction model in the embodiment of the present disclosure is as follows: Figure 1-Figure 4 The power bidding prediction model optimization method shown in the embodiment is generated.

[0127] Thus, through Figure 1-Figure 4 The target power bidding prediction model generated by the power bidding prediction model optimization method shown in the embodiment predicts the usage forecast value, generates the power bidding value of the target object, and predicts the user's power consumption through the LSTM model. Starting from the top-level design of the power market, with the goal of minimizing the risk of power system operation and maximizing the suppression of user load uncertainty, it is considered to combine heuristic algorithms to solve constrained optimization problems in order to make the results more robust. It is considered to combine deep learning to predict user power consumption based on historical portraits of user power consumption, so that it meets the robust bidding model of the power sales company, so as to cope with the system errors in medium and long-term power trading contracts due to the use of planned values ​​for power, improve the operating profit and market competitiveness of the power sales company, and reduce economic risks.

[0128] Corresponding to the electricity bidding prediction model optimization method provided in the above-mentioned embodiments, an embodiment of the present disclosure also provides an electricity bidding prediction model optimization device. Since the electricity bidding prediction model optimization device provided in the embodiment of the present disclosure corresponds to the electricity bidding prediction model optimization method provided in the above-mentioned embodiments, the implementation method of the above-mentioned electricity bidding prediction model optimization method is also applicable to the electricity bidding prediction model optimization device provided in the embodiment of the present disclosure, and will not be described in detail in the following embodiments.

[0129] Figure 6 is a schematic diagram of an electricity bidding prediction model optimization device according to an embodiment of the present disclosure, such as Figure 6 As shown, the power bidding prediction model optimization device 600 includes: an acquisition module 610, a parsing module 620, a building module 630 and a generation module 640.

[0130] The acquisition module 610 is used to acquire the unit configuration data and user historical load information of the target object.

[0131] The analysis module 620 is used to analyze the user's historical load information and unit configuration data to determine the target factor.

[0132] The establishment module 630 is used to establish an initial power bidding prediction model for the target object based on the target factors and the preset balance constraints.

[0133] The generation module 640 is used to iteratively adjust the model parameters of the initial power bidding prediction model based on the firefly algorithm until a preset end condition is met, thereby generating a target power bidding prediction model.

[0134] According to one embodiment of the present disclosure, historical user load information and unit configuration data are parsed to determine a target factor, including: semantically processing historical user load information to generate multiple candidate factors; calculating contribution values ​​of multiple candidate factors to electricity bidding; and determining a target factor from multiple candidate factors based on the contribution values.

[0135] According to one embodiment of the present disclosure, an initial electricity bidding prediction model of a target object is established based on target factors and preset balance constraints, including: establishing an objective function of the target object based on the target factors; establishing an initial electricity bidding prediction model of the target object based on the objective function and balance constraints.

[0136] According to one embodiment of the present disclosure, the target factors include the actual electricity sales to users, the actual electricity sales price to users, the long-term contract electricity price, the electricity purchase price from the power market, the electricity purchase volume from the power market, and the monthly decomposition value of the annual long-term contract electricity purchased by the power sales company. The target function is: Among them, t is the bidding time interval, η t To actually sell electricity to users, R i,t The actual electricity price sold to users, M contact,t is the monthly decomposition value of the annual long-term contract electricity purchased by the power sales company, C contact,t is the long-term contract electricity price, C buy,t is the price of electricity purchased from the electricity market, M buy,t To purchase electricity from the electricity market, maxP total Profit data for the target object.

[0137] According to an embodiment of the present disclosure, the balancing constraint condition includes an electricity balancing constraint condition, an electricity purchase constraint condition and a quotation condition.

[0138] According to one embodiment of the present disclosure, the model parameters of the initial electricity bidding prediction model are iteratively adjusted based on the firefly algorithm until a preset end condition is met to generate a target electricity bidding prediction model, including: determining the model parameter range of the initial electricity bidding prediction model, and randomly generating an initial firefly factor set, wherein the position of any candidate firefly factor in the firefly factor set represents a set of model parameters of the initial electricity bidding prediction model; iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm; repeating the above process of iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm until the end condition is met to generate the target electricity bidding prediction model.

[0139] Therefore, the method of this scheme can improve the accuracy of electricity bidding prediction, and can be used as a basis for optimizing resource allocation, while enhancing the competitiveness of users in the electricity market.

[0140] Corresponding to the electricity bidding prediction method provided in the above-mentioned embodiments, an embodiment of the present disclosure also provides an electricity bidding prediction device. Since the electricity bidding prediction device provided in the embodiment of the present disclosure corresponds to the electricity bidding prediction method provided in the above-mentioned embodiments, the implementation method of the above-mentioned electricity bidding prediction method is also applicable to the electricity bidding prediction device provided in the embodiment of the present disclosure, and will not be described in detail in the following embodiments.

[0141] Figure 7 is a schematic diagram of a power bidding prediction device according to an embodiment of the present disclosure. Figure 7 As shown, the power bidding prediction device 700 includes: a prediction module 710 and an input module 720.

[0142] The prediction module 710 is used to predict the user's power consumption through the user's power consumption model to generate a usage prediction value, wherein the user's power consumption model is an LSTM model.

[0143] The input module 720 is used to input the usage prediction value into the target power bidding prediction model to generate the power bidding value of the target object, wherein the target power bidding prediction model is as follows: Figure 1-Figure 4 The electricity bidding prediction model optimization method shown in the embodiment is generated.

[0144] Thus, through Figure 1-Figure 4 The target power bidding prediction model generated by the power bidding prediction model optimization method shown in the embodiment predicts the usage forecast value, generates the power bidding value of the target object, and predicts the user's power consumption through the LSTM model. Starting from the top-level design of the power market, with the goal of minimizing the risk of power system operation and maximizing the suppression of user load uncertainty, it is considered to combine heuristic algorithms to solve constrained optimization problems in order to make the results more robust. It is considered to combine deep learning to predict user power consumption based on historical portraits of user power consumption, so that it meets the robust bidding model of the power sales company, so as to cope with the system errors in medium and long-term power trading contracts due to the use of planned values ​​for power, improve the operating profit and market competitiveness of the power sales company, and reduce economic risks.

[0145] In order to implement the above embodiment, the present disclosure further provides an electronic device 800, Figure 8 is a schematic diagram of an electronic device according to an embodiment of the present disclosure, such as Figure 8As shown, the electronic device 800 includes: a processor 801 and a memory 802 that is communicatively connected to the processor, the memory 802 stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor 801 to implement the present disclosure. Figure 1-Figure 4 The power bidding prediction model optimization method of the embodiment, or Figure 5 A method for predicting power bidding according to an embodiment of the present invention.

[0146] In order to implement the above embodiments, the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to implement the above embodiments. Figure 1-Figure 4 The power bidding prediction model optimization method of the embodiment, or Figure 5 A method for predicting power bidding according to an embodiment of the present invention.

[0147] In order to implement the above embodiments, the present disclosure also provides a computer program product, including a computer program. When the computer program is executed by a processor, the computer program implements the above embodiments. Figure 1-Figure 4 The power bidding prediction model optimization method of the embodiment, or Figure 5 A method for predicting power bidding according to an embodiment of the present invention.

[0148] It should be noted that personal information from users should be collected for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. In addition, such collection / sharing should be carried out after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign the agreement / authorization including authorization of relevant user information before the user uses the function. In addition, any necessary steps should be taken to protect and safeguard access to such personal information data and ensure that others who have access to personal information data comply with its privacy policy and procedures.

[0149] The present application is expected to provide an implementation scheme for users to selectively block the use or access of personal information data. That is, the present disclosure is expected to provide hardware and / or software to prevent or block access to such personal information data. Once the personal information data is no longer needed, the risk can be minimized by limiting data collection and deleting the data. In addition, when applicable, such personal information is de-identified to protect the privacy of the user.

[0150] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0151] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0152] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0153] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that contains, stores, communicates, propagates or transmits a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0154] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0155] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0156] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0157] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for optimizing an electricity bidding prediction model, characterized in that: include: Obtain the unit configuration data and user historical load information of the target object; Analyzing the user's historical load information and the unit configuration data to determine the target factor; Establishing an initial power bidding prediction model for the target object based on the target factor and preset balance constraints; The model parameters of the initial power bidding prediction model are iteratively adjusted based on the firefly algorithm until a preset end condition is met, thereby generating a target power bidding prediction model.

2. The method according to claim 1, characterized in that The parsing of the user's historical load information and the unit configuration data to determine the target factor includes: Performing semantic processing on the user historical load information to generate multiple candidate factors; Calculate the contribution of multiple candidate factors to electricity bidding; The target factor is determined from a plurality of candidate factors based on the contribution value.

3. The method according to claim 1 or 2, characterized in that: The initial power bidding prediction model of the target object is established based on the target factor and the preset balance constraint condition, including: Establishing the objective function of the target object based on the target factor; An initial electricity bidding prediction model for the target object is established based on the objective function and the balance constraint condition.

4. The method according to claim 3, characterized in that The target factors include actual electricity sales to users, actual electricity sales price to users, long-term contract electricity price, electricity purchase price from the power market, electricity purchase volume from the power market, and monthly decomposition value of annual long-term contract electricity purchased by power sales companies. The target function is: Wherein, t is the bidding time interval, and n t is the actual amount of electricity sold to the user, and R i,t is the actual electricity price sold to users, and M contact,t is the monthly decomposition value of the annual long-term contract electricity volume of the power sales company, and the C contact,t is the long-term contract electricity price, and C buy,t is the price of electricity purchased from the power market, and M buy,t is the amount of electricity purchased from the electricity market, the maxP total Profit data for the target object.

5. The method according to claim 3, characterized in that: The balance constraint conditions include power balance constraint conditions, power purchase constraint conditions and quotation conditions.

6. The method according to claim 1, characterized in that The method of iteratively adjusting the model parameters of the initial power bidding prediction model based on the firefly algorithm until a preset end condition is met to generate a target power bidding prediction model includes: Determine a model parameter range of the initial power bidding prediction model, and randomly generate an initial firefly factor set, wherein the position of any candidate firefly factor in the firefly factor set represents a set of model parameters of the initial power bidding prediction model; Iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm, and determining whether the updated candidate firefly factors meet the end condition; The above process of iteratively updating the positions of the candidate firefly factors in the firefly factor set based on the firefly algorithm is repeated until the end condition is met, thereby generating a target electricity bidding prediction model.

7. A method for predicting power bidding, characterized in that: include: Predicting the user's power consumption through a user power consumption model to generate a usage prediction value, wherein the user power consumption model is an LSTM model; The usage forecast value is input into a target power bidding forecast model to generate a power bidding value of a target object, wherein the target power bidding forecast model is generated by the power bidding forecast model optimization method according to any one of claims 1-6.

8. An electricity bidding prediction model optimization device, characterized in that: include: An acquisition module is used to obtain the unit configuration data and user historical load information of the target object; An analysis module, used for analyzing the user's historical load information and the unit configuration data to determine the target factor; An establishment module, used to establish an initial power bidding prediction model of the target object based on the target factor and the preset balance constraint condition; The generation module is used to iteratively adjust the model parameters of the initial power bidding prediction model based on the firefly algorithm until a preset end condition is met, thereby generating a target power bidding prediction model.

9. An electric power bidding prediction device, characterized in that: include: A prediction module, used to predict the user's power consumption through a user's power consumption model to generate a usage prediction value, wherein the user's power consumption model is an LSTM model; An input module is used to input the usage forecast value into a target power bidding forecast model to generate a power bidding value of a target object, wherein the target power bidding forecast model is generated by the power bidding forecast model optimization method according to any one of claims 1-6.

10. An electronic device, characterized in that: Including memory. processor; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the power bidding prediction model optimization method as described in any one of claims 1 to 6, or the power bidding prediction method as described in claim 7.

Images