Auxiliary decision-making method, device and equipment for power transaction, storage medium and product

Abstract

The invention discloses an auxiliary decision-making method, device and equipment for power transaction, a storage medium and a product, and relates to the technical field of power spot transaction. The method comprises the following steps: carrying out random simulation on predicted source data of power transaction in a preset time period based on a Monte Carlo model to obtain transaction strategy probability models under different transaction strategies; randomly generating a group of initial transaction strategies as an initial population, and inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain prediction transaction data of each individual in the initial population; the initial population is selected, crossed and varied based on the predicted transaction data of each individual, an optimal transaction strategy is obtained, and the optimal transaction strategy is used for aiding decision-making of the power transaction, so that the transaction prediction accuracy can be improved, an aided decision-making method with reference value is provided for a trader, and the trader is helped to make a transaction strategy in advance.

Description

Technical Field

[0001] The present application relates to the technical field of electricity spot trading, and in particular to an auxiliary decision-making method, device, equipment, storage medium and product for electricity trading. Background Art

[0002] With the continuous growth of electricity demand and the advancement of energy structure adjustment, electricity spot trading, as a flexible and efficient way to allocate electricity resources, has gradually become an important means to promote reform in the power sector. Spot trading allows market participants to buy and sell electricity at a near-real-time price based on real-time supply and demand conditions and superior dispatch, thereby improving the efficiency and economy of the power market.

[0003] However, existing electricity trading systems generally have a limited ability for trading models to capture dynamic changes in the market. The idealized assumptions in the design of trading models are inconsistent with the complexity of the actual electricity market, resulting in a large deviation between the model prediction results and the actual market situation. Summary of the invention

[0004] The present application provides an auxiliary decision-making method, device, storage medium and product for power trading to solve the problem that the prediction results of the existing power trading model are not accurate, cannot reflect the complex changes in the actual power market, and are difficult to help traders formulate trading strategies in advance.

[0005] In a first aspect, an embodiment of the present application provides an auxiliary decision-making method for power trading, comprising:

[0006] Based on the Monte Carlo model, the forecast source data of power trading within the preset time period is randomly simulated to obtain the trading strategy probability model under different trading strategies;

[0007] Randomly generate a set of initial trading strategies as an initial population, and input the trading strategy probability model and the prediction source data into a preset deep learning model to obtain predicted trading data for each individual in the initial population;

[0008] The initial population is selected, crossed and mutated based on the predicted transaction data of each of the individuals to obtain an optimal transaction strategy; the optimal transaction strategy is used to assist in decision-making on power transactions.

[0009] In a second aspect, an embodiment of the present application provides an auxiliary decision-making method for power trading, including:

[0010] A strategy probability model acquisition module is used to perform random simulation on the forecast source data of power transactions within a preset time period based on a Monte Carlo model to obtain a trading strategy probability model under different trading strategies;

[0011] A transaction data prediction module, used to randomly generate a set of initial transaction strategies as an initial population, and input the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain predicted transaction data for each individual in the initial population;

[0012] The trading strategy optimization module is used to select, cross and mutate the initial population based on the predicted trading data of each individual to obtain the optimal trading strategy; the optimal trading strategy is used to assist in decision-making on power trading.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising:

[0014] at least one processor;

[0015] and a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the auxiliary decision-making method for power trading described in any embodiment of the present application.

[0017] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the auxiliary decision-making method for power trading described in any embodiment of the present application when executed.

[0018] In a fifth aspect, an embodiment of the present application provides a computer program product including a computer program, which, when executed by a processor, implements the auxiliary decision-making method for power trading described in any embodiment of the present application.

[0019] The technical solution of the embodiment of the present application is to obtain a trading strategy probability model under different trading strategies by randomly simulating the predicted source data of power transactions within a preset time period based on a Monte Carlo model; randomly generate a group of initial trading strategies as an initial population, and input the trading strategy probability model and the predicted source data into a preset deep learning model to obtain the predicted trading data of each individual in the initial population; select, crossover and mutate the initial population based on the predicted trading data of each individual to obtain the optimal trading strategy, which is used to assist in decision-making on power transactions, improve the accuracy of trading predictions, provide trading strategies with reference value, provide auxiliary information for traders' trading decisions, and help traders formulate trading strategies in advance, thereby solving the problem that the prediction results of existing power trading models are not accurate, cannot reflect the complex changes in the actual power market, and are difficult to help traders formulate trading strategies in advance.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 A flowchart of an auxiliary decision-making method for power trading provided in Example 1 of the present application;

[0023] Figure 2 A flowchart of an auxiliary decision-making method for power trading provided in Embodiment 2 of the present application;

[0024] Figure 3 A schematic diagram of the structure of an auxiliary decision-making device for power trading provided in Embodiment 3 of the present application;

[0025] Figure 4 It is a structural diagram of an electronic device that implements the auxiliary decision-making method for power trading in an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0028] Figure 1This is a flowchart of a method for assisting decision-making in power trading provided in an embodiment of the present application. This embodiment can be applied to situations where intelligent models are used to assist traders in formulating power trading strategies. This method can be performed by an auxiliary decision-making device for power trading. The auxiliary decision-making device for power trading can be implemented in the form of hardware and / or software. The auxiliary decision-making device for power trading can be configured in an electronic device. Figure 1 As shown, the method includes:

[0029] S110, performing random simulation on the forecast source data of power trading within a preset time period based on the Monte Carlo model to obtain a trading strategy probability model under different trading strategies.

[0030] Among them, the forecast source data can be understood as the data source needed for power trading forecasting. Generally, the forecast source data comes from the data disclosed by the power grid and the power trading impact data released by the power trading related departments. The forecast source data at least includes: spot trading data, power grid data and trading constraint data.

[0031] The Monte Carlo method is a random method based on sampling and the generation of random numbers, so it is also called random sampling method, computer random simulation method, etc. The basic principle of the Monte Carlo method is to obtain the probability of a certain event occurring as an approximate solution to the problem through digital simulation experiments. The trading strategy can be understood as the trading strategy in the power market simulated by the Monte Carlo model. The trading strategy probability model can be understood as the probability model of the change of trading data over time under different trading strategies within a preset time period.

[0032] Specifically, the raw data of power transactions within a preset time period are obtained from multiple data sources. In order to further improve the accuracy of transaction prediction, after the raw data is classified, the data is processed according to different categories to obtain high-quality prediction source data. Data preprocessing may include: data cleaning, data format unification, and data analysis and calculation. In order to facilitate data storage and reading, the prediction source data can also be stored in the database of each category according to the data category for use.

[0033] The prediction source data is input into the Monte Carlo model, and the prediction source data is randomly simulated through the Monte Carlo model to generate the probability distribution of the simulated trading strategy composed of the prediction source data under different values. The trading strategy probability model is constructed based on the simulated trading strategy and the corresponding probability distribution.

[0034] This step ensures the timeliness and accuracy of information by collecting massive data from the power trading market and power plants in real time; and through comprehensive preprocessing technology, it can effectively clean noise data and reasonably fill in missing values ​​to improve data quality.

[0035] S120. Randomly generate a set of initial trading strategies as an initial population, and input the trading strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted trading data of each individual in the initial population.

[0036] Among them, the initial population can be understood as the population initially established in the genetic algorithm. The population is a cluster composed of multiple individuals. In this embodiment, each trading strategy is an individual, so the initial population is a cluster of multiple trading strategies. The preset deep learning model can be understood as a deep learning model deployed in an electronic device, which can be a general deep learning model or a complete deep learning model obtained by training with transaction data training samples, and is used to generate predicted transaction data. The predicted transaction data can be understood as the predicted transaction data, for example, it can include: predicted revenue, predicted cost, predicted output and predicted power.

[0037] Specifically, multiple trading strategies are constructed according to the trading information that the trading strategies need to contain, and each trading strategy is taken as an individual to form the initial population. The trading strategy probability model and the predicted source data are input into the preset deep learning model, and after the prediction of the deep learning model, the predicted trading data of each individual in the initial population, that is, each trading strategy, is obtained.

[0038] In an optional embodiment, before the step of inputting the trading strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted trading data of each individual in the initial population, it also includes: obtaining a strategy evaluation index; calculating the strategy evaluation value of each trading strategy in the trading strategy probability model on the strategy evaluation index based on the prediction source data; and discarding the trading strategies whose strategy evaluation values ​​in the trading strategy probability model do not meet the preset evaluation conditions.

[0039] Among them, the strategy evaluation index can be understood as an index for evaluating the performance of the trading strategy, for example, it can be trading output or power. The strategy evaluation index can be set according to the actual needs of the trader, and this embodiment does not impose any restrictions on this.

[0040] Specifically, after obtaining the trading strategy probability model, for each trading strategy in the trading strategy probability model, the strategy evaluation value of the trading strategy on the preset strategy evaluation indicator is calculated according to the prediction source data. If the strategy evaluation value does not meet the preset evaluation conditions, such as the trading output is higher than the output threshold, the trading strategy that does not meet the preset evaluation conditions will be discarded.

[0041] This embodiment aims to delete trading strategies that obviously do not meet the actual trading needs of traders or exceed the trading capabilities of traders before searching for the optimal solution for the trading strategies, thereby reducing the amount of calculation for subsequent optimal solution searching.

[0042] S130. Select, crossover and mutate the initial population based on the predicted transaction data of each individual to obtain an optimal transaction strategy, which is used to assist in decision-making on power transactions.

[0043] Among them, the optimal trading strategy can be understood as the trading strategy with the best performance in one or more indicators obtained from the evaluation of simulated trading strategies.

[0044] Specifically, a genetic algorithm is used to select, cross and mutate individuals in the initial population according to the predicted transaction data to obtain the optimal transaction strategy, which can help traders make decisions in the power trading market.

[0045] In this embodiment, the deep learning algorithm and the genetic algorithm are combined to effectively improve the accuracy of transaction prediction and transaction strategy selection, so that the predicted optimal transaction strategy can adapt to the complex changes in the power market.

[0046] The technical solution of the embodiment of the present application is to obtain a trading strategy probability model under different trading strategies by randomly simulating the predicted source data of power trading within a preset time period based on a Monte Carlo model; randomly generate a group of initial trading strategies as an initial population, and input the trading strategy probability model and the predicted source data into a preset deep learning model to obtain the predicted trading data of each individual in the initial population; select, crossover and mutate the initial population based on the predicted trading data of each individual to obtain the optimal trading strategy, which is used to assist in decision-making in power trading, can improve the accuracy of trading predictions, provide trading strategies with reference value, provide auxiliary information for traders' trading decisions, and help traders formulate trading strategies in advance.

[0047] Figure 2 A flowchart of another auxiliary decision-making method for power trading provided in an embodiment of the present application, based on the above embodiment, this embodiment further refines the prediction step of transaction data. The prediction source data includes at least: transaction settlement data, power grid data and transaction constraint data; accordingly, the step of inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population includes: for each individual in the initial population, determining the power bidding space according to the power grid data, and determining the price control coefficient according to the transaction constraint data and the transaction settlement data; inputting the power bidding space, the price control coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual.

[0048] like Figure 2 As shown, the method includes:

[0049] S210, performing random simulation on the forecast source data of power trading within a preset time period based on the Monte Carlo model to obtain a trading strategy probability model under different trading strategies.

[0050] Among them, the forecast source data at least includes: transaction settlement data, power grid data and transaction constraint data. Spot transaction data can be understood as the spot transaction settlement data of the power trading market on the day before, for example, it can include: unified clearing results, spot clearing electricity price, real-time inter-provincial spot clearing electricity price and real-time inter-provincial spot clearing electricity quantity. Power grid data can be understood as the power generation plan data of the power station, for example, it can include: positive and negative standby demand data on the day before, power transmission and transformation equipment maintenance plan, must-start and must-stop units, day-ahead interconnection line plan, day-ahead new energy output. Transaction constraint data can be understood as data constrained by power transactions, which can be derived from power transaction policy constraints, such as the upper limit of transaction electricity price. The preset time period can be a few days before the transaction day. Since the power transaction policies of different provinces or regions are different, the forecast source data within the corresponding time period can be obtained according to the actual power transaction forecast demand. In addition, since weather conditions will directly cause changes in power demand, the forecast source data can also include: meteorological data. Meteorological data refers to weather condition data on the day before the transaction, which can include: weather conditions, such as cloudy, sunny or rainy and snowy days, as well as weather temperature.

[0051] S220. For each individual in the initial population, determine the power bidding space according to the power grid data, and determine the price control coefficient according to the transaction constraint data and the transaction settlement data.

[0052] Among them, the power bidding space can be understood as a certain proportion of the total market demand that needs to be traded through bidding according to market rules. The price control coefficient can be understood as a coefficient used to control the power transaction price. The price control coefficient is subject to transaction constraints and the market transaction average price and transaction day average price in the transaction settlement data.

[0053] Specifically, the key factors affecting the electricity bidding space include: benchmark electricity price, market demand influencing factors and power generation costs. The benchmark electricity price is usually determined by the government or regulatory agency based on the average cost and reasonable profit of the electricity market as the basis for the electricity bidding space. Market demand and power generation costs are key factors affecting the electricity bidding space. Power generation companies need to calculate in detail various costs including fuel costs, maintenance costs, depreciation costs, etc. to obtain the total power generation cost, analyze the supply and demand conditions in the market, predict the trend of electricity prices, and thus determine a reasonable quotation range. Therefore, the electricity bidding space = benchmark electricity price ± (market demand influencing factor × power generation cost). Among them, the market demand influencing factor is a coefficient determined based on the market supply and demand conditions, which reflects the degree of market influence on electricity prices.

[0054] According to the key factors affecting the power bidding space, it is further deduced that the power bidding space is determined according to the sum of the day-ahead interconnection line plan, day-ahead new energy output, the whole network load and the output of non-market units contained in the power grid data. Exemplarily, the power bidding space may include: day-ahead power bidding space and real-time power bidding space; day-ahead power bidding space = predicted whole network load + abs (day-ahead interconnection line plan) - day-ahead new energy output - non-market unit output; actual power bidding space = actual whole network load + abs (day-ahead interconnection line plan) - day-ahead new energy output - non-market unit output.

[0055] If the average price on the trading day is greater than the average market price, the price control coefficient is equal to the average market price divided by the average price on the trading day; otherwise, the price control coefficient is 1. That is, when the average price on the trading day is greater than the average market price, the average price on the trading day is controlled.

[0056] S230: Input the power bidding space, price control coefficient, transaction settlement data, power grid data and transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual.

[0057] Specifically, the electricity bidding space, price control coefficient, transaction settlement data, power grid data and transaction strategy probability model are input into the preset deep learning model. The deep learning model predicts and executes the transaction strategy represented by each individual to generate predicted transaction data.

[0058] Optionally, the predicted transaction data may include: predicted revenue, predicted cost, predicted price, predicted output and predicted power, reflecting the benefits and risks brought by the trading strategy from multiple aspects.

[0059] In an optional embodiment, the step of inputting the power bidding space, the control coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual includes:

[0060] For each individual, inputting the power bidding space, the transaction settlement data, and the power grid data into the output prediction module of the deep learning model to obtain the predicted output of the individual;

[0061] Inputting the power bidding space, the transaction settlement data and the power grid data into the power prediction module of the deep learning model to obtain the predicted power of the individual;

[0062] Inputting the predicted output, the predicted power, the transaction settlement data and the power grid data into the cost prediction module of the deep learning model to obtain the predicted cost of the individual;

[0063] Inputting the price control coefficient, the transaction settlement data and the power grid data into the price prediction module of the deep learning model to obtain the predicted price of the individual;

[0064] The predicted output, the predicted power, the predicted price and the predicted cost are input into the profit prediction module of the deep learning model to obtain the predicted profit of the individual.

[0065] Among them, the preset deep learning model includes: output prediction module, power prediction module, cost prediction module, price prediction module and income prediction module. The output prediction module is used to predict the output under the trading strategy, the power prediction module is used to predict the power under the trading strategy, and the cost prediction module is used to predict the cost under the trading strategy. The price prediction module is used to predict the price under the trading strategy, and the income prediction module is used to predict the income under the trading strategy.

[0066] Specifically, for each individual trading strategy in the initial population, the power bidding space, transaction settlement data and power grid data under the trading strategy are input into the output prediction module of the preset deep learning model to obtain the predicted output output by the output prediction module. Then the power bidding space, transaction settlement data and power grid data are input into the power prediction module of the deep learning model to obtain the predicted power. After obtaining the predicted output and predicted power, they are input into the cost prediction module of the deep learning model together with the transaction settlement data and power grid data to obtain the predicted cost. The price control coefficient, transaction settlement data and power grid data are input into the price prediction module of the deep learning model to obtain the predicted price. And the obtained predicted output, predicted power, predicted price and predicted cost are input into the revenue prediction module of the deep learning model to obtain the predicted revenue.

[0067] The higher the predicted output, predicted power or predicted cost, the higher the risk and cost of the trading strategy; the higher the predicted price and predicted return, the greater the gain of the trading strategy. The performance of the trading strategy can be estimated comprehensively based on the predicted output, predicted power, predicted cost and predicted return as well as the type of trading strategy.

[0068] It is understandable that the specific calculation rules for predicted output, predicted power, predicted cost and predicted revenue in different regions, time periods or power trading markets may be different, and the embodiments of the present invention are not limited to this.

[0069] S240. Select, crossover and mutate the initial population based on the predicted transaction data of each individual to obtain an optimal transaction strategy; the optimal transaction strategy is used to assist in decision-making on power transactions.

[0070] In an optional embodiment, the optimal trading strategy is intuitively presented to the trader through charts or reports. In addition, the actual trading strategy of the trading day can be obtained, and the actual trading strategy and the optimal trading strategy can be analyzed and compared to obtain the review results. The review results can provide traders with strategy evaluation and optimization suggestions to help traders improve their trading success rate.

[0071] The technical solution of the embodiment of the present application is to obtain a trading strategy probability model under different trading strategies by randomly simulating the predicted source data of power transactions within a preset time period based on a Monte Carlo model; for each individual in the initial population, the power bidding space is determined according to the power grid data, and the price control coefficient is determined according to the transaction constraint data and the transaction settlement data; the power bidding space, the price control coefficient, the transaction settlement data, the power grid data and the trading strategy probability model are input into a preset deep learning model to obtain the predicted trading data of each individual; based on the predicted trading data of each individual, the initial population is selected, crossed and mutated to obtain the optimal trading strategy; the optimal trading strategy is used to assist in decision-making on power transactions, which can improve the accuracy of transaction predictions, provide trading strategies with reference value, provide auxiliary information for traders' trading decisions, and help traders formulate trading strategies in advance.

[0072] In an optional embodiment, the initial population is selected, crossed and mutated based on the predicted transaction data of each of the individuals, and the step of obtaining the optimal trading strategy includes: calculating the fitness of each of the individuals according to the expected trading strategy type and the predicted transaction data; performing selection operations, crossover operations and mutation operations according to the fitness to create new individuals, and inheriting them to the new generation of population; returning to execute the step of calculating the fitness of each of the individuals according to the expected trading strategy type and the predicted transaction data; until the termination condition is reached, the individual with the maximum fitness obtained in the evolution process is determined as the optimal trading strategy.

[0073] Among them, fitness can be understood as an evaluation index for selecting individuals. In this embodiment, the calculation function of fitness is related to the expected trading strategy type and can be designed according to the expected trading strategy. The expected trading strategy type can be understood as the type of trading strategy that the trader hopes to implement. Optionally, the expected trading strategy includes: aggressive trading strategy, conservative trading strategy and balanced trading strategy.

[0074] Specifically, a genetic algorithm is used to calculate the fitness according to the expected trading strategy type and the predicted trading data of each individual, sort the individuals according to the fitness, select the individual with the largest fitness for crossover and mutation to create a new individual, and pass it on to the new generation of population. For the new generation of population, the step of calculating the fitness of each individual according to the expected trading strategy type and the predicted trading data is returned until the termination condition is reached, and the individual with the largest fitness obtained in the evolution process is determined as the optimal trading strategy. The termination condition can be that the number of iterations reaches the preset number, or the fitness meets the requirements.

[0075] Optionally, the step of calculating the fitness of each of the individuals based on the expected trading strategy type and the predicted trading data includes: for an aggressive trading strategy, calculating the fitness of each of the individuals based on the predicted trading data and a first fitness function; the first fitness function is the predicted profit; for a conservative trading strategy, calculating the fitness of each of the individuals based on the predicted trading data and a second fitness function; the second fitness function is the total predicted cost; the total predicted cost is a negative number; for a balanced trading strategy, calculating the fitness of each of the individuals based on the predicted trading data and a third fitness function; the third fitness function is the weighted sum of the predicted profit and the total predicted cost.

[0076] Among them, the weights of the predicted revenue and the total predicted cost in the third fitness function can be set according to actual business needs, and this embodiment does not impose any restrictions on this.

[0077] Specifically, the fitness function corresponding to the expected trading strategy is selected to calculate the fitness of the individual, and the individual with excellent performance is selected from the population as the evaluation indicator of the individual performance. The corresponding trading strategy can be selected according to the trader's preference for the expected trading strategy, so that the optimal trading strategy finally obtained is more in line with the needs of the trader.

[0078] Figure 3 This is a schematic diagram of the structure of an auxiliary decision-making device for power trading provided in an embodiment of the present application. Figure 3 As shown, the device includes: a strategy probability model acquisition module 310, a transaction data prediction module 320 and a transaction strategy optimization module 330; wherein,

[0079] A strategy probability model acquisition module 310 is used to perform random simulation on the prediction source data of power trading within a preset time period based on a Monte Carlo model to obtain a trading strategy probability model under different trading strategies;

[0080] The transaction data prediction module 320 is used to randomly generate a set of initial transaction strategies as an initial population, and input the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population;

[0081] The trading strategy optimization module 330 is used to select, cross and mutate the initial population based on the predicted trading data of each of the individuals to obtain the optimal trading strategy; the optimal trading strategy is used to assist in decision-making on power trading.

[0082] The technical solution of the embodiment of the present application is to obtain a trading strategy probability model under different trading strategies by randomly simulating the predicted source data of power trading within a preset time period based on a Monte Carlo model; randomly generate a group of initial trading strategies as an initial population, and input the trading strategy probability model and the predicted source data into a preset deep learning model to obtain the predicted trading data of each individual in the initial population; select, crossover and mutate the initial population based on the predicted trading data of each individual to obtain the optimal trading strategy, which is used to assist in decision-making in power trading, can improve the accuracy of trading predictions, provide trading strategies with reference value, provide auxiliary information for traders' trading decisions, and help traders formulate trading strategies in advance.

[0083] Optionally, the prediction source data includes at least: transaction settlement data, power grid data and transaction constraint data.

[0084] Optionally, the transaction data prediction module 320 includes:

[0085] a determining unit, configured to determine, for each individual of the initial population, a power bidding space according to the power grid data, and determine a price regulation coefficient according to the transaction constraint data and the transaction settlement data;

[0086] A model prediction unit is used to input the electricity bidding space, the price control coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual.

[0087] Optionally, the predicted transaction data includes: predicted revenue, predicted cost, predicted price, predicted output and predicted power.

[0088] The model prediction unit is specifically used for:

[0089] For each individual, inputting the power bidding space, the transaction settlement data, and the power grid data into the output prediction module of the deep learning model to obtain the predicted output of the individual;

[0090] Inputting the power bidding space, the transaction settlement data and the power grid data into the power prediction module of the deep learning model to obtain the predicted power of the individual;

[0091] Inputting the predicted output, the predicted power, the transaction settlement data and the power grid data into the cost prediction module of the deep learning model to obtain the predicted cost of the individual;

[0092] Inputting the price control coefficient, the transaction settlement data and the power grid data into the price prediction module of the deep learning model to obtain the predicted price of the individual;

[0093] The predicted output, the predicted power, the predicted price and the predicted cost are input into the profit prediction module of the deep learning model to obtain the predicted profit of the individual.

[0094] Optionally, the trading strategy optimization module 330 includes:

[0095] A fitness calculation unit, used to calculate the fitness of each individual according to the expected trading strategy type and the predicted trading data;

[0096] Individual genetic units, used to perform selection operations, crossover operations and mutation operations according to the fitness to create new individuals, and pass them on to the new generation of populations;

[0097] A return execution unit, used for returning to execute the step of calculating the fitness of each individual according to the expected trading strategy type and the predicted trading data;

[0098] The trading strategy selection unit is used to determine the individual with the maximum fitness obtained in the evolution process as the optimal trading strategy until the termination condition is reached.

[0099] Optional, fitness calculation unit, specifically used for:

[0100] For an aggressive trading strategy, the fitness of each of the individuals is calculated according to the predicted trading data and a first fitness function; the first fitness function is the predicted return;

[0101] For a conservative trading strategy, the fitness of each individual is calculated according to the predicted trading data and a second fitness function; the second fitness function is a predicted cost; and the predicted cost is a negative number;

[0102] For the balanced trading strategy, the fitness of each of the individuals is calculated based on the predicted trading data and a third fitness function; the second fitness function is a weighted sum of predicted benefits and predicted costs.

[0103] Optionally, the device further comprises:

[0104] An indicator acquisition module, used for acquiring a strategy evaluation indicator before the step of inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population;

[0105] An indicator calculation module, used for calculating the strategy evaluation value of each of the trading strategies in the trading strategy probability model on the strategy evaluation indicator according to the prediction source data;

[0106] The strategy abandonment module is used to abandon the trading strategies whose strategy evaluation values ​​in the trading strategy probability model do not meet the preset evaluation conditions.

[0107] The auxiliary decision-making device for power trading provided in the embodiments of the present application can execute the auxiliary decision-making method for power trading provided in any embodiment of the present application, and has the corresponding functional modules and beneficial effects of the execution method.

[0108] Figure 4 A schematic diagram of an electronic device 10 that can be used to implement an embodiment of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0109] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0110] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0111] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as an auxiliary decision-making method for power trading.

[0112] In some embodiments, the auxiliary decision-making method for power trading can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the auxiliary decision-making method for power trading described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the auxiliary decision-making method for power trading in any other appropriate manner (for example, by means of firmware).

[0113] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0114] In some embodiments, the auxiliary decision-making method for power trading can be implemented as a computer program, which is invisibly included in a computer program product. When the computer program is executed by a processor, the auxiliary decision-making method for power trading of the present application is implemented. The computer program product can be understood as a software product that mainly implements its solution through a computer program. The computer program for implementing the method of the present application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the computer program, when executed by the processor, implements the functions / operations specified in the flowchart and / or block diagram. The computer program can be executed entirely on the machine, partially on the machine, partially on the machine as an independent software package and partially on a remote machine, or entirely on a remote machine or server.

[0115] In the context of the present application, a computer readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device or equipment. A computer readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium may be a machine readable signal medium. A more specific example of a machine readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0116] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0117] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0118] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0119] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution of this application can be achieved, and this document is not limited here.

[0120] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. An auxiliary decision-making method for power trading, characterized in that: include: Based on the Monte Carlo model, the forecast source data of power trading within the preset time period is randomly simulated to obtain the trading strategy probability model under different trading strategies; Randomly generate a set of initial trading strategies as an initial population, and input the trading strategy probability model and the prediction source data into a preset deep learning model to obtain predicted trading data for each individual in the initial population; The initial population is selected, crossed and mutated based on the predicted transaction data of each of the individuals to obtain an optimal transaction strategy; the optimal transaction strategy is used to assist in decision-making on power transactions.

2. The auxiliary decision-making method for power trading according to claim 1, characterized in that: The prediction source data at least includes: transaction settlement data, power grid data and transaction constraint data; accordingly, the step of inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population includes: For each individual in the initial population, determining the power bidding space according to the power grid data, and determining the price regulation coefficient according to the transaction constraint data and the transaction settlement data; The electricity bidding space, the price control coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model are input into a preset deep learning model to obtain the predicted transaction data of each individual.

3. The auxiliary decision-making method for power trading according to claim 2, characterized in that: The predicted transaction data includes: predicted revenue, predicted cost, predicted price, predicted output and predicted power; the step of inputting the power bidding space, the regulation coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual includes: For each individual, inputting the power bidding space, the transaction settlement data, and the power grid data into the output prediction module of the deep learning model to obtain the predicted output of the individual; Inputting the power bidding space, the transaction settlement data and the power grid data into the power prediction module of the deep learning model to obtain the predicted power of the individual; Inputting the predicted output, the predicted power, the transaction settlement data and the power grid data into the cost prediction module of the deep learning model to obtain the predicted cost of the individual; Inputting the price control coefficient, the transaction settlement data and the power grid data into the price prediction module of the deep learning model to obtain the predicted price of the individual; The predicted output, the predicted power, the predicted price and the predicted cost are input into the profit prediction module of the deep learning model to obtain the predicted profit of the individual.

4. The auxiliary decision-making method for power trading according to claim 1, characterized in that: The step of selecting, crossing over and mutating the initial population based on the predicted transaction data of each individual to obtain the optimal transaction strategy comprises: Calculating the fitness of each of the individuals according to the expected trading strategy type and the predicted trading data; Perform selection, crossover and mutation operations according to the fitness to create new individuals, and pass them on to the new generation of population; Return to execute the step of calculating the fitness of each individual according to the expected trading strategy type and the predicted trading data; Until the termination condition is reached, the individual with the maximum fitness obtained in the evolution process is determined as the optimal trading strategy.

5. The auxiliary decision-making method for power trading according to claim 4, characterized in that: The step of calculating the fitness of each individual according to the expected trading strategy type and the predicted trading data comprises: For an aggressive trading strategy, the fitness of each of the individuals is calculated according to the predicted trading data and a first fitness function; the first fitness function is the predicted return; For a conservative trading strategy, the fitness of each individual is calculated according to the predicted trading data and a second fitness function; the second fitness function is a predicted cost; and the predicted cost is a negative number; For the balanced trading strategy, the fitness of each of the individuals is calculated based on the predicted trading data and a third fitness function; the third fitness function is a weighted sum of the predicted benefit and the predicted cost.

6. The auxiliary decision-making method for power trading according to claim 1, characterized in that: Before the step of inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population, the step further includes: Get strategy evaluation metrics; Calculate the strategy evaluation value of each of the trading strategies in the trading strategy probability model on the strategy evaluation index according to the prediction source data; The trading strategies whose strategy evaluation values ​​in the trading strategy probability model do not meet the preset evaluation conditions are discarded.

7. An auxiliary decision-making device for power trading, characterized in that: include: A strategy probability model acquisition module is used to perform random simulation on the forecast source data of power transactions within a preset time period based on a Monte Carlo model to obtain a trading strategy probability model under different trading strategies; A transaction data prediction module, used to randomly generate a set of initial transaction strategies as an initial population, and input the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain predicted transaction data for each individual in the initial population; The trading strategy optimization module is used to select, cross and mutate the initial population based on the predicted trading data of each individual to obtain the optimal trading strategy; the optimal trading strategy is used to assist in decision-making on power trading.

8. The auxiliary decision-making device for power trading according to claim 7, characterized in that: The prediction source data at least includes: transaction settlement data, power grid data and transaction constraint data; The transaction data prediction module includes: a determining unit, configured to determine, for each individual of the initial population, a power bidding space according to the power grid data, and determine a price regulation coefficient according to the transaction constraint data and the transaction settlement data; A model prediction unit is used to input the electricity bidding space, the price control coefficient, the transaction settlement data, the power grid data and the transaction strategy probability model into a preset deep learning model to obtain the predicted transaction data of each individual.

9. The auxiliary decision-making device for power trading according to claim 7, characterized in that: The trading strategy optimization module includes: A fitness calculation unit, used to calculate the fitness of each individual according to the expected trading strategy type and the predicted trading data; Individual genetic units, used to perform selection operations, crossover operations and mutation operations according to the fitness to create new individuals, and pass them on to the new generation of populations; A return execution unit, used for returning to execute the step of calculating the fitness of each individual according to the expected trading strategy type and the predicted trading data; The trading strategy selection unit is used to determine the individual with the maximum fitness obtained in the evolution process as the optimal trading strategy until the termination condition is reached.

10. The auxiliary decision-making device for power trading according to claim 7, characterized in that: The device also includes: An indicator acquisition module, used for acquiring a strategy evaluation indicator before the step of inputting the transaction strategy probability model and the prediction source data into a preset deep learning model to obtain the predicted transaction data of each individual in the initial population; An indicator calculation module, used for calculating the strategy evaluation value of each of the trading strategies in the trading strategy probability model on the strategy evaluation indicator according to the prediction source data; The strategy abandonment module is used to abandon the trading strategies whose strategy evaluation values ​​in the trading strategy probability model do not meet the preset evaluation conditions.

11. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the auxiliary decision-making method for electricity trading according to any one of claims 1-6.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the auxiliary decision-making method for power trading according to any one of claims 1 to 6 when executed.

13. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the auxiliary decision-making method for power trading according to any one of claims 1-6.

Images