Energy storage resource allocation method, device, medium and equipment based on power Internet of Things

Abstract

The present invention relates to a method, device, medium and equipment for allocating energy storage resources based on the electric power Internet of Things. The method predicts the state of charge of the energy storage terminal to obtain a reference state of charge sequence of the energy storage terminal, determines the energy storage terminal's power adjustment interval in combination with the state of charge limit of the energy storage terminal, reconstructs the predicted load sequence of the power grid, and uses a training loss function for supervision to generate a reference load sequence that meets the expected peak shaving and valley filling. A reference correction sequence is determined based on the predicted load sequence and the reference load sequence, and the reference correction sequence is evaluated based on the reference correction sequence and the power adjustment interval. When the evaluation result meets the preset conditions, it indicates that the energy storage terminal can meet the required power adjustment, thereby improving the accuracy of energy storage resource allocation, ensuring that the expected effect of peak shaving and valley filling can be achieved, and maximizing the expected effect of peak shaving and valley filling through iterative training of the load reconstruction model.

Description

Technical Field

[0001] The present invention is applicable to the field of electric power technology, and in particular relates to a method, device, medium and equipment for allocating energy storage resources based on the electric power Internet of Things. Background Art

[0002] In power applications, peak shaving and valley filling refers to the process of storing energy during low-load periods through energy storage and releasing it during peak-load periods, thereby balancing power supply and demand and optimizing grid operations. The effectiveness of peak shaving and valley filling can typically be characterized by the peak-valley difference, which refers to the difference between the grid's maximum and minimum loads before and after peak shaving and valley filling.

[0003] Due to the limited capacity of a single energy storage terminal, the existing technology usually responds to peak shaving and valley filling by coordinating multiple energy storage terminals. However, for electrochemical energy storage terminals, when the energy storage terminal is over-discharged or over-charged, it will accelerate battery aging and even cause safety issues. Therefore, when performing peak shaving and valley filling, the energy storage terminal will find it difficult to meet its estimated adjusted power due to the influence of the battery status of the energy storage terminal. In other words, the accuracy of energy storage resource allocation is poor, which makes the peak shaving and valley filling effect difficult to achieve the expected effect.

[0004] Therefore, how to improve the accuracy of energy storage resource allocation and maximize the expected effect of peak shaving and valley filling while ensuring that the expected effect of peak shaving and valley filling can be achieved has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method, apparatus, medium, and device for allocating energy storage resources based on the power Internet of Things to solve the problem.

[0006] In a first aspect, a method for allocating energy storage resources based on the power Internet of Things is provided, the method comprising:

[0007] Obtain the historical state of charge sequences corresponding to M energy storage terminals, where M is a positive integer;

[0008] For any energy storage terminal, a reference state of charge sequence corresponding to the energy storage terminal is obtained based on the historical state of charge sequence corresponding to the energy storage terminal and the trained state of charge prediction model;

[0009] Determine a dischargeable amount sequence and a chargeable amount sequence corresponding to the energy storage end according to a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, and a maximum amount of charge corresponding to the energy storage end;

[0010] Determine the overall discharge capacity sequence based on the discharge capacity sequences corresponding to the M energy storage terminals;

[0011] Determine the overall charge capacity sequence based on the discharge capacity sequences corresponding to the M energy storage terminals;

[0012] According to the acquired historical load sequence and the trained load forecasting model, the forecast load sequence is obtained;

[0013] Obtaining a reference load sequence according to the predicted load sequence, the sequence reconstruction model and the training loss function;

[0014] Determine a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determine a reference correction value sequence based on the reference load difference sequence;

[0015] determining an evaluation parameter of the reference correction amount sequence according to the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence;

[0016] If the evaluation parameter satisfies a preset condition, determining the reference correction value sequence as the target correction value sequence;

[0017] According to the target correction amount sequence, the estimated charge amount or the estimated discharge amount corresponding to each of the M energy storage terminals at a preset time point is determined.

[0018] In a second aspect, a device for allocating energy storage resources based on the power Internet of Things is provided, the device comprising:

[0019] A state of charge acquisition module is used to obtain the historical state of charge sequences corresponding to M energy storage terminals, where M is a positive integer;

[0020] The state of charge prediction module is used to obtain a reference state of charge sequence corresponding to any energy storage terminal based on the historical state of charge sequence corresponding to the energy storage terminal and the trained state of charge prediction model;

[0021] A power determination module is used to determine a dischargeable power sequence and a chargeable power sequence corresponding to the energy storage end based on a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, a state of charge lower limit value corresponding to the energy storage end, and a maximum power corresponding to the energy storage end;

[0022] A first sequence determination module is used to determine an overall discharge capacity sequence based on the dischargeable capacity sequences corresponding to the M energy storage terminals;

[0023] A second sequence determination module is used to determine the overall charge capacity sequence according to the dischargeable capacity sequences corresponding to the M energy storage terminals;

[0024] The load forecasting module is used to obtain the predicted load sequence based on the acquired historical load sequence and the trained load forecasting model;

[0025] A sequence reconstruction module, configured to obtain a reference load sequence based on the predicted load sequence, the sequence reconstruction model, and the training loss function;

[0026] A correction value determination module is used to determine a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determine a reference correction value sequence based on the reference load difference sequence;

[0027] a sequence evaluation module, configured to determine an evaluation parameter of the reference correction amount sequence based on the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence;

[0028] a first condition judgment module, configured to determine the reference correction value sequence as a target correction value sequence if the evaluation parameter satisfies a preset condition;

[0029] The power distribution module is used to determine the estimated charge capacity or estimated discharge capacity corresponding to M energy storage terminals at preset time points according to the target correction amount sequence.

[0030] In a third aspect, an embodiment of the present invention provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the energy storage resource allocation method based on the power Internet of Things as described in the first aspect is implemented.

[0031] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the energy storage resource allocation method based on the power Internet of Things as described in the first aspect is implemented.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] By predicting the state of charge of the energy storage end, the reference state of charge sequence of the energy storage end is obtained. The energy storage end's power adjustment range is determined in combination with the state of charge limit of the energy storage end. The predicted load sequence of the power grid is reconstructed and supervised by a training loss function to generate a reference load sequence that meets the expected peak shaving and valley filling. The reference correction sequence is determined based on the predicted load sequence and the reference load sequence. The reference correction sequence is evaluated based on the reference correction sequence and the power adjustment range. When the evaluation result meets the preset conditions, it indicates that the energy storage end can meet the required power adjustment, thereby improving the accuracy of energy storage resource allocation and ensuring that the expected effect of peak shaving and valley filling can be achieved. The expected effect of peak shaving and valley filling is maximized through iterative training of the load reconstruction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 This is a schematic diagram of an application environment of an energy storage resource allocation method based on the power Internet of Things provided in the first embodiment of the present invention;

[0036] Figure 2 This is a flow chart of a method for allocating energy storage resources based on the power Internet of Things provided in the first embodiment of the present invention;

[0037] Figure 3 This is a structural diagram of an energy storage resource allocation device based on the power Internet of Things provided in the second embodiment of the present invention;

[0038] Figure 4 This is a structural diagram of a computer device for a method for allocating energy storage resources based on the power Internet of Things provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0039] In the following description, specific details such as particular system structures and techniques are provided for purposes of illustration, not limitation, to facilitate a thorough understanding of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

[0040] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0041] It will also be understood that the term "and / or" used in the present description and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0042] As used in the present specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0043] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0044] References to "one embodiment" or "some embodiments" in the present specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present invention. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in yet other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0045] It should be understood that the order of execution of the steps in the following embodiments does not necessarily mean the order in which they are executed. The order in which each process is executed should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0046] In order to illustrate the technical solution of the present invention, specific embodiments are provided below.

[0047] The first embodiment of the present invention provides a method for allocating energy storage resources based on the power Internet of Things, which can be applied in the following situations: Figure 1 In an application environment, a server communicates with a client. Clients include, but are not limited to, PDAs, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, cloud-based devices, and personal digital assistants (PDAs). The server can be implemented as a standalone server or a server cluster consisting of multiple servers.

[0048] See also Figure 2, is a flow chart of a method for allocating energy storage resources based on the power Internet of Things provided by the first embodiment of the present invention. The above energy storage resource allocation method can be applied to Figure 1 The server in the example can obtain the state of charge corresponding to each energy storage terminal through the power Internet of Things, that is, the sensors deployed at each energy storage terminal. After the sensors at each energy storage terminal collect the state of charge locally, they transmit the collected state of charge to the server in a preset communication method. The server is deployed with a trained state of charge prediction model, a trained load prediction model, and a sequence reconstruction model. Figure 2 As shown, the energy storage resource allocation method may include the following steps:

[0049] Step S201: Obtain historical state of charge sequences corresponding to M energy storage terminals.

[0050] Wherein, M is a positive integer, the energy storage end may refer to an electrochemical energy storage type energy storage end, the historical state of charge sequence may include historical states of charge corresponding to K historical time points, K is a positive integer, and the time interval between two adjacent historical time points may be a fixed value T. In this embodiment, T may be set to 5 minutes, and the implementer may adjust the value of T according to actual conditions.

[0051] Specifically, for any energy storage terminal, the historical state of charge of the energy storage terminal can be collected by the state of charge sensor deployed at the energy storage terminal at the corresponding historical time point. The state of charge sensor of each energy storage terminal belongs to the power Internet of Things, and each energy storage terminal can communicate with the service end to transmit the collected state of charge.

[0052] Step S202 : For any energy storage terminal, a reference state of charge sequence corresponding to the energy storage terminal is obtained based on a historical state of charge sequence corresponding to the energy storage terminal and a trained state of charge prediction model.

[0053] Among them, the trained state of charge prediction model can adopt a time series prediction model, and the time series prediction model can use a recurrent neural network (RNN) model, a long short-term memory network (LSTM) model, a temporal convolutional network (TCN) model, etc.

[0054] Specifically, this embodiment is described using the TCN model as an example. The sliding window length corresponding to the TCN model is L. Then, for any energy storage end, based on the K-L+1th historical state of charge to the Kth historical state of charge in the historical state of charge sequence corresponding to the energy storage end, the reference state of charge at the first preset time point can be predicted. Then, based on the K-L+2th historical state of charge to the Kth historical state of charge in the historical state of charge sequence corresponding to the energy storage end and the reference state of charge at the first preset time point, the reference state of charge at the second preset time point can be predicted, and so on, until the reference state of charge at the Nth preset time point is obtained.

[0055] It should be noted that, in order to ensure the prediction accuracy of the trained state of charge prediction model, the time interval between the Kth historical time point and the first preset time point is T, and the time interval between two adjacent preset time points is also T.

[0056] The training process of the state of charge prediction model may include, for any energy storage end, predicting the sample state of charge at the L+1th historical time point based on the 1st historical state of charge to the Lth historical state of charge in the historical state of charge sequence corresponding to the energy storage end, and calculating the state of charge prediction model sub-loss based on the historical state of charge at the L+1th historical time point and the sample state of charge at the L+1th historical time point in combination with the mean square error loss function. Similarly, based on the 2nd historical state of charge to the L+1th historical state of charge in the historical state of charge sequence corresponding to the energy storage end, predicting the sample state of charge at the L+2th historical time point. According to the historical state of charge at the L+2th historical time point and the sample state of charge at the L+2th historical time point, combined with the mean square error loss function, the state of charge prediction model sub-loss can also be calculated. Similarly, a single energy storage terminal can correspond to KL state of charge prediction model sub-losses, and M energy storage terminals can correspond to M*(KL) state of charge prediction model sub-losses. The state of charge prediction model loss is formed by all the state of charge prediction model sub-losses. According to the state of charge prediction model loss, the state of charge prediction model is trained using the gradient descent method until the state of charge prediction model loss converges, and the trained state of charge prediction model can be obtained.

[0057] Step S203 , determining a dischargeable capacity sequence and a chargeable capacity sequence corresponding to the energy storage end according to a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, a state of charge lower limit value corresponding to the energy storage end, and a maximum capacity corresponding to the energy storage end.

[0058] Among them, the upper limit value of the state of charge can be used to prevent battery overcharging, and the lower limit value of the state of charge can be used to prevent deep discharge of the battery. The basic upper limit value of the state of charge and the basic lower limit value of the state of charge can be determined locally at the energy storage end through charge and discharge testing methods, accelerated aging tests, etc. Since this embodiment is applied to the peak shaving and valley filling scenario, the implementer can adjust the basic upper limit value of the state of charge and the basic lower limit value of the state of charge, for example, by increasing the basic upper limit value of the state of charge and lowering the basic lower limit value of the state of charge.

[0059] The maximum power may refer to the maximum power capacity of the energy storage end.

[0060] The dischargeable amount sequence may include the dischargeable amounts corresponding to N preset time points. The dischargeable amount may refer to the maximum discharge amount of the energy storage end at the preset time point under the premise that the charge state upper limit value and charge state lower limit value corresponding to the energy storage end are met.

[0061] The chargeable capacity sequence may include the chargeable capacities corresponding to N preset time points. The chargeable capacity may refer to the maximum charge capacity of the energy storage end at the preset time point, provided that the upper limit value and the lower limit value of the state of charge corresponding to the energy storage end are met.

[0062] Optionally, the reference state of charge sequence includes reference states of charge corresponding to N preset time points, where N is a positive integer;

[0063] Determining a dischargeable capacity sequence and a chargeable capacity sequence corresponding to the energy storage end according to a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, and a maximum capacity corresponding to the energy storage end, including:

[0064] For any preset time point, determine a first state of charge difference based on the state of charge upper limit and the reference state of charge corresponding to the preset time point, and determine the dischargeable capacity of the energy storage end corresponding to the preset time point based on the first state of charge difference and the maximum capacity;

[0065] Determine a second state of charge difference based on the reference state of charge and the lower limit of the state of charge corresponding to the preset time point, and determine the chargeable capacity of the energy storage end corresponding to the preset time point based on the second state of charge difference and the maximum capacity;

[0066] Traversing N preset time points, the dischargeable amounts corresponding to the obtained N preset time points form a dischargeable amount sequence corresponding to the energy storage end, and the chargeable amounts corresponding to the obtained N preset time points form a chargeable amount sequence corresponding to the energy storage end.

[0067] Among them, the first state of charge difference can be obtained by subtracting the state of charge upper limit value from the reference state of charge corresponding to the preset time point. By multiplying the first state of charge difference and the maximum power, the dischargeable capacity of the energy storage end corresponding to the preset time point can be obtained.

[0068] The second state of charge difference can be obtained by subtracting the reference state of charge corresponding to the preset time point from the lower limit of the state of charge. The second state of charge difference and the maximum power can be multiplied to obtain the chargeable capacity of the energy storage end corresponding to the preset time point.

[0069] Specifically, the discharge amounts corresponding to the N preset time points are sequentially arranged to form a discharge amount sequence corresponding to the energy storage end, and the chargeable amounts corresponding to the N preset time points are sequentially arranged to form a chargeable amount sequence corresponding to the energy storage end.

[0070] Step S204 : determining an overall discharge capacity sequence according to the dischargeable capacity sequences corresponding to the M energy storage terminals.

[0071] The overall discharge amount sequence may include overall discharge amounts corresponding to N preset time points.

[0072] Specifically, for any preset time point, the dischargeable capacities corresponding to the M energy storage terminals at the preset time point are added together to obtain the overall discharge capacity corresponding to the preset time point. By traversing N preset time points, the overall discharge capacities corresponding to the N preset time points are obtained. The overall discharge capacities corresponding to the N preset time points are used to form an overall discharge capacity sequence in chronological order.

[0073] Step S205 : determining an overall charge capacity sequence according to the dischargeable capacity sequences corresponding to the M energy storage terminals.

[0074] The overall charge amount sequence may include overall charge amounts corresponding to N preset time points.

[0075] Specifically, for any preset time point, the chargeable capacities corresponding to the M energy storage terminals at the preset time point are added together to obtain the overall charge capacity corresponding to the preset time point. By traversing N preset time points, the overall charge capacities corresponding to the N preset time points are obtained, and the overall charge capacity sequence is formed by the overall charge capacities corresponding to the N preset time points in chronological order.

[0076] Step S206: Obtain a predicted load sequence based on the acquired historical load sequence and the trained load prediction model.

[0077] Among them, the historical load sequence can include the power grid loads corresponding to K historical time points respectively. The trained load forecasting model can also adopt the time series forecasting model. Therefore, the reasoning and training process of the load forecasting model will not be described in detail here.

[0078] The predicted load sequence may include predicted loads corresponding to N preset time points.

[0079] Step S207: Obtain a reference load sequence according to the predicted load sequence, the sequence reconstruction model and the training loss function.

[0080] Among them, the sequence reconstruction model can be used to reconstruct the predicted load sequence to output the expected load sequence after peak shaving and valley filling, that is, the reference load sequence.

[0081] Specifically, the sequence reconstruction model adopts an online learning method. During the use of the sequence reconstruction model, the sequence reconstruction model is trained in real time according to the training loss function.

[0082] Optionally, a reference load sequence is obtained based on the predicted load sequence, the sequence reconstruction model, and the training loss, including:

[0083] Input the predicted load sequence into the sequence reconstruction model to obtain the reconstructed load sequence;

[0084] The training loss is calculated based on the reconstructed load sequence, the predicted load sequence and the training loss function;

[0085] Update the parameters of the sequence reconstruction model according to the training loss until the training loss converges to obtain a trained sequence reconstruction model;

[0086] The predicted load sequence is input into the trained sequence reconstruction model to obtain the reference load sequence.

[0087] The reconstructed load sequence may refer to an intermediate output result of a sequence reconstruction model, and the training loss may be used for training the sequence reconstruction model.

[0088] Specifically, after the sequence reconstruction model is trained according to the training loss and a trained sequence reconstruction model is obtained, the output of the trained sequence reconstruction model after the predicted load sequence is inputted is used as a reference load sequence.

[0089] Optionally, a training loss is calculated based on the reconstructed load sequence, the predicted load sequence, and the training loss function, including:

[0090] Accumulating the predicted load sequence to obtain a first load amount;

[0091] Accumulating the reconstructed load sequence to obtain a second load amount;

[0092] determining a first sub-loss according to a difference between the first load amount and the second load amount;

[0093] The variance of the reconstructed load sequence is used as the second sub-loss;

[0094] Multiplying the first sub-loss by the preset first weight to obtain a first multiplication result;

[0095] Multiplying the second sub-loss by the preset second weight to obtain a second multiplication result;

[0096] The sum of the first multiplication result and the second multiplication result is used as the training loss.

[0097] The first load amount may be used to represent the predicted total load amount within a time period formed by N preset time points, and the second load amount may be used to represent the reconstructed total load amount within a time period formed by N preset time points.

[0098] Specifically, the absolute value of the difference between the first load and the second load is used as the first sub-loss. The first sub-loss can be used to monitor that the total load of the power grid remains as consistent as possible before and after peak shaving and valley filling.

[0099] The variance of the reconstructed load sequence can characterize the volatility of the reconstructed load sequence, and can also characterize the effect after peak shaving and valley filling. The smaller the variance of the reconstructed load sequence, the smaller the peak-valley difference is, and the better the effect after peak shaving and valley filling is.

[0100] The preset first weight can be used to control the influence of the first sub-loss on the training loss. In order to ensure that the total load of the power grid is stable before and after peak shaving and valley filling, and thus provide stable power supply to the power consumption side, the implementer can set the preset first weight to a larger value so that the first sub-loss is prioritized to be smaller during the training process.

[0101] The preset second weight can be used to control the influence of the second sub-loss on the training loss, and the initial value of the preset second weight can be 1.

[0102] Step S208: determining a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determining a reference correction value sequence based on the reference load difference sequence.

[0103] Among them, for any preset time point, the predicted load corresponding to the preset time point in the predicted load sequence is subtracted from the reference load corresponding to the preset time point in the reference load sequence, and the subtraction result is used as the reference load difference corresponding to the preset time point. N preset time points are traversed to obtain the reference load difference corresponding to the N preset time points respectively. The reference load difference sequence is formed in chronological order by the reference load difference corresponding to the N preset time points respectively. It can be known that the reference load difference can be a negative value.

[0104] Specifically, the accumulated value of the reference load difference corresponding to the first preset time point to the i-th preset time point is used as the reference correction value for the i-th preset time point, where i is an integer in the range of [1, N]. By traversing the values of i, the reference correction values corresponding to the N preset time points can be obtained, and the reference correction values corresponding to the N preset time points form a reference correction value sequence in chronological order.

[0105] Step S209 : determining evaluation parameters of the reference correction amount sequence according to the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence.

[0106] The evaluation parameter can be used to characterize the degree of adaptability of the reference correction sequence to the energy storage terminal state.

[0107] Optionally, the reference correction amount sequence includes reference correction amounts corresponding to N preset time points, the overall discharge amount sequence includes overall discharge amounts corresponding to N preset time points, and the overall charge amount sequence includes overall charge amounts corresponding to N preset time points, where N is a positive integer.

[0108] According to the reference correction amount sequence, the overall discharge amount sequence and the overall charge amount sequence, the evaluation parameters of the reference correction amount sequence are determined, including:

[0109] For any preset time point, if the reference correction amount corresponding to the preset time point is less than the overall discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point is greater than the overall charge amount corresponding to the preset time point, determine the sub-evaluation value corresponding to the preset time point as the first preset value;

[0110] Otherwise, calculating a first difference between the total discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point;

[0111] Calculating a second difference between the reference correction amount corresponding to the preset time point and the overall charge amount corresponding to the preset time point;

[0112] The smaller value of the first difference and the second difference is used as the target value, the target value is mapped according to the reference correction amount corresponding to the preset time point, and the mapping result is used as the sub-evaluation value corresponding to the preset time point;

[0113] Traverse N preset time points, add up the sub-evaluation values corresponding to the N preset time points, and use the added result as the evaluation parameter of the reference correction value sequence.

[0114] The first preset value may be 0, and both the first difference and the second difference are processed by taking their absolute values.

[0115] Specifically, the target value is compared with the reference correction amount corresponding to the preset time point, and the ratio calculation result is used as the mapping processing result.

[0116] It can be seen that the larger the evaluation parameter, the worse the adaptability of the reference correction sequence to the energy storage end state, and the smaller the evaluation parameter, the better the adaptability of the reference correction sequence to the energy storage end state. The minimum evaluation parameter is 0.

[0117] In step S210 , if the evaluation parameters meet the preset conditions, the reference correction amount sequence is determined as the target correction amount sequence.

[0118] The preset conditions may be used to evaluate whether the power adjustment of the reference correction amount sequence can be achieved without overcharging or overdischarging each energy storage terminal.

[0119] Specifically, when the evaluation parameters meet the preset conditions, it can be considered that the power adjustment corresponding to the reference correction amount sequence can be achieved without overcharging or overdischarging at each energy storage end, and the reference correction amount sequence is determined to be the target correction amount sequence.

[0120] Optionally, after determining the evaluation parameters of the reference correction value sequence, the method further includes:

[0121] If the evaluation parameter does not meet the preset conditions, the second weight is updated with a preset step size, and the process returns to the step of obtaining a reference load sequence based on the predicted load sequence, the sequence reconstruction model, and the training loss function.

[0122] Among them, when the evaluation parameters do not meet the preset conditions, it can be considered that the power adjustment corresponding to the reference correction amount sequence cannot be achieved under the condition that each energy storage end is not overcharged or over-discharged. At this time, it means that the peak shaving and valley filling effect of the reference load sequence is difficult to meet, so the reference load sequence needs to be updated.

[0123] Specifically, in this embodiment, the preset step size can be -0.05, that is, when the evaluation parameter does not meet the preset conditions, the second weight is reduced, so that the supervision of the second sub-loss is reduced when the reference load sequence is generated. In other words, the peak shaving and valley filling effect represented by the reference load sequence output by the sequence reconstruction model can be appropriately reduced, thereby adapting to the battery status of the energy storage end.

[0124] Step S211 : determining the estimated charge amount or the estimated discharge amount corresponding to each of the M energy storage terminals at a preset time point according to the target correction amount sequence.

[0125] Among them, the estimated charging amount and the estimated discharging amount can be used to assist in formulating peak shaving and valley filling plans for each preset time point. The peak shaving and valley filling plans can be used to transmit to each energy storage terminal so that the energy storage terminal can perform charging and discharging control at the preset time point based on its own situation, or provide feedback to the service terminal in advance when the control requirements cannot be met.

[0126] Optionally, the target correction amount sequence includes target correction amounts corresponding to N preset time points, the dischargeable amount sequence includes dischargeable amounts corresponding to N preset time points, and the chargeable amount sequence includes chargeable amounts corresponding to N preset time points, where N is a positive integer.

[0127] Determine the estimated charge or discharge amounts corresponding to the M energy storage terminals at the preset time points according to the target correction amount sequence, including:

[0128] For any preset time point, if the target correction amount corresponding to the preset time point is less than the second preset value, then the estimated charge capacity corresponding to each of the M energy storage terminals is determined based on the target correction amount corresponding to the preset time point and the chargeable capacity of each of the M energy storage terminals at the preset time point;

[0129] If the target correction amount corresponding to the preset time point is greater than or equal to the second preset value, the estimated discharge amounts corresponding to the M energy storage ends are determined according to the target correction amount corresponding to the preset time point and the dischargeable amounts corresponding to the M energy storage ends at the preset time point.

[0130] Among them, for any energy storage end, the chargeable amount of the energy storage end at the preset time point is compared with the overall charge amount corresponding to the preset time point to obtain the charge amount ratio of the energy storage end at the preset time point, and the target correction amount corresponding to the preset time point and the charge amount ratio of the energy storage end at the preset time point are multiplied to obtain the estimated charge amount of the energy storage end at the preset time point.

[0131] Similarly, for any energy storage end, the dischargeable amount of the energy storage end at the preset time point is compared with the overall discharge amount corresponding to the preset time point to obtain the discharge amount ratio of the energy storage end at the preset time point. The target correction amount corresponding to the preset time point and the discharge amount ratio of the energy storage end at the preset time point are multiplied to obtain the estimated discharge amount of the energy storage end at the preset time point.

[0132] In this embodiment, the state of charge of the energy storage terminal is predicted to obtain a reference state of charge sequence of the energy storage terminal. The energy storage terminal's power adjustment interval is determined in combination with the state of charge limit of the energy storage terminal. The predicted load sequence of the power grid is reconstructed and supervised by a training loss function to generate a reference load sequence that meets the expected peak shaving and valley filling. A reference correction value sequence is determined based on the predicted load sequence and the reference load sequence. The reference correction value sequence is evaluated based on the reference correction value sequence and the power adjustment interval. When the evaluation result meets the preset conditions, it indicates that the energy storage terminal can meet the required power adjustment, thereby improving the accuracy of energy storage resource allocation, ensuring that the expected effect of peak shaving and valley filling can be achieved, and maximizing the expected effect of peak shaving and valley filling through iterative training of the load reconstruction model.

[0133] Corresponding to the method of the above embodiment, Figure 3 A schematic diagram of the structure of an energy storage resource allocation device based on the Power Internet of Things (PoI), provided in accordance with a second embodiment of the present invention, is shown. This device is applied to a server, which can obtain the state of charge (SOC) corresponding to each energy storage terminal through the PoI, or sensors deployed at each energy storage terminal. After locally collecting the SOC, the sensors at each energy storage terminal transmit the collected SOC to the server via a preset communication method. The server is equipped with a trained SOC prediction model, a trained load prediction model, and a sequence reconstruction model. For ease of illustration, only the portions relevant to this embodiment of the present invention are shown.

[0134] See also Figure 3 , the energy storage resource allocation device includes:

[0135] The state of charge acquisition module 301 is used to obtain the historical state of charge sequences corresponding to M energy storage terminals, where M is a positive integer;

[0136] The state of charge prediction module 302 is used to obtain a reference state of charge sequence corresponding to any energy storage terminal based on the historical state of charge sequence corresponding to the energy storage terminal and the trained state of charge prediction model;

[0137] The power determination module 303 is used to determine the dischargeable power sequence and the chargeable power sequence corresponding to the energy storage end based on the reference state of charge sequence corresponding to the energy storage end, the state of charge upper limit value corresponding to the energy storage end, the state of charge lower limit value corresponding to the energy storage end, and the maximum power corresponding to the energy storage end;

[0138] A first sequence determination module 304 is configured to determine an overall discharge capacity sequence based on the dischargeable capacity sequences corresponding to the M energy storage terminals;

[0139] A second sequence determination module 305 is configured to determine an overall charge capacity sequence based on the dischargeable capacity sequences corresponding to the M energy storage terminals;

[0140] The load forecasting module 306 is used to obtain a forecast load sequence based on the acquired historical load sequence and the trained load forecasting model;

[0141] A sequence reconstruction module 307 is used to obtain a reference load sequence based on the predicted load sequence, the sequence reconstruction model and the training loss function;

[0142] The correction amount determination module 308 is used to determine a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determine a reference correction amount sequence based on the reference load difference sequence;

[0143] A sequence evaluation module 309 is configured to determine an evaluation parameter of a reference correction amount sequence based on the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence;

[0144] A first condition judgment module 310 is configured to determine the reference correction value sequence as the target correction value sequence if the evaluation parameter satisfies a preset condition;

[0145] The power distribution module 311 is used to determine the estimated charge capacity or estimated discharge capacity corresponding to the M energy storage terminals at preset time points according to the target correction amount sequence.

[0146] Optionally, the reference state of charge sequence includes reference states of charge corresponding to N preset time points, where N is a positive integer;

[0147] The power determination module 303 includes:

[0148] a dischargeable capacity determination unit, configured to determine, for any preset time point, a first state of charge difference based on the state of charge upper limit and a reference state of charge corresponding to the preset time point, and determine the dischargeable capacity of the energy storage terminal corresponding to the preset time point based on the first state of charge difference and the maximum capacity;

[0149] a chargeable capacity determination unit, configured to determine a second state-of-charge difference based on a reference state-of-charge and a lower limit of the state-of-charge corresponding to the preset time point, and determine the chargeable capacity of the energy storage terminal corresponding to the preset time point based on the second state-of-charge difference and the maximum capacity;

[0150] The sequence forming unit is used to traverse N preset time points, and form a dischargeable amount sequence corresponding to the energy storage end from the dischargeable amounts corresponding to the N preset time points, and form a chargeable amount sequence corresponding to the energy storage end from the chargeable amounts corresponding to the N preset time points.

[0151] Optionally, the sequence reconstruction module 307 includes:

[0152] A model reconstruction unit is used to input the predicted load sequence into the sequence reconstruction model to obtain a reconstructed load sequence;

[0153] A loss calculation unit, configured to calculate the training loss based on the reconstructed load sequence, the predicted load sequence, and the training loss function;

[0154] The model training unit is used to update the parameters of the sequence reconstruction model according to the training loss until the training loss converges to obtain a trained sequence reconstruction model;

[0155] The sequence output unit is used to input the predicted load sequence into the trained sequence reconstruction model to obtain the reference load sequence.

[0156] Optionally, the loss calculation unit includes:

[0157] A first load calculation subunit is configured to accumulate the predicted load sequence to obtain a first load;

[0158] A second load calculation subunit is configured to accumulate the reconstructed load sequence to obtain a second load;

[0159] a first sub-loss calculation subunit, configured to determine a first sub-loss according to a difference between the first load amount and the second load amount;

[0160] a second sub-loss calculation subunit, configured to use the variance of the reconstructed load sequence as the second sub-loss;

[0161] A first weighting subunit, configured to multiply the first sub-loss by a preset first weight to obtain a first multiplication result;

[0162] A second weighting subunit is used to multiply the second sub-loss by a preset second weight to obtain a second multiplication result;

[0163] The loss determination subunit is configured to use the sum of the first multiplication result and the second multiplication result as the training loss.

[0164] Optionally, the energy storage resource allocation device further includes:

[0165] The second condition judgment module is used to update the second weight with a preset step size if the evaluation parameter does not meet the preset condition, and return to execute the step of obtaining a reference load sequence based on the predicted load sequence, the sequence reconstruction model and the training loss function.

[0166] Optionally, the reference correction amount sequence includes reference correction amounts corresponding to N preset time points, the overall discharge amount sequence includes overall discharge amounts corresponding to N preset time points, and the overall charge amount sequence includes overall charge amounts corresponding to N preset time points, where N is a positive integer.

[0167] The sequence evaluation module 309 includes:

[0168] a first sub-evaluation value calculation unit, configured to, for any preset time point, determine that the sub-evaluation value corresponding to the preset time point is a first preset value if the reference correction amount corresponding to the preset time point is less than the overall discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point is greater than the overall charge amount corresponding to the preset time point;

[0169] a first difference calculation unit, configured to calculate a first difference between the overall discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point;

[0170] a second difference calculation unit, configured to calculate a second difference between the reference correction amount corresponding to the preset time point and the overall charge amount corresponding to the preset time point;

[0171] a second sub-evaluation value calculation unit, configured to use the smaller value of the first difference and the second difference as a target value, perform mapping processing on the target value according to the reference correction amount corresponding to the preset time point, and use the mapping processing result as the sub-evaluation value corresponding to the preset time point;

[0172] The evaluation parameter determination unit is used to traverse N preset time points, add the sub-evaluation values corresponding to the N preset time points, and use the addition result as the evaluation parameter of the reference correction value sequence.

[0173] Optionally, the target correction amount sequence includes target correction amounts corresponding to N preset time points, the dischargeable amount sequence includes dischargeable amounts corresponding to N preset time points, and the chargeable amount sequence includes chargeable amounts corresponding to N preset time points, where N is a positive integer.

[0174] The power distribution module 311 includes:

[0175] a charge capacity allocation unit, configured to, for any preset time point, determine an estimated charge capacity corresponding to each of the M energy storage terminals based on the target correction amount corresponding to the preset time point and the chargeable capacities of the M energy storage terminals corresponding to the preset time point, if the target correction amount corresponding to the preset time point is less than a second preset value;

[0176] The discharge capacity allocation unit is configured to determine the estimated discharge capacities corresponding to the M energy storage terminals respectively according to the target correction amount corresponding to the preset time point and the dischargeable capacities of the M energy storage terminals respectively at the preset time point, if the target correction amount corresponding to the preset time point is greater than or equal to the second preset value.

[0177] It should be noted that the information interaction, execution process, etc. between the above-mentioned modules, units, and sub-units are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0178] Figure 4 This is a schematic diagram of the structure of a computer device for a method of allocating energy storage resources based on the power Internet of Things provided in the third embodiment of the present invention. Figure 4 As shown, the computer device of this embodiment includes: at least one processor ( Figure 4 Only one is shown), a memory, and a computer program stored in the memory and executable on at least one processor. When the processor executes the computer program, the steps in any of the above-mentioned embodiments of the energy storage resource allocation method based on the power Internet of Things are implemented.

[0179] The computer device may include, but is not limited to, a processor and a memory. It will be understood by those skilled in the art that Figure 4 The above is merely an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include a network interface, a display screen, and an input device.

[0180] The processor may be a CPU, other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0181] Memory includes readable storage media, internal memory, and the like. Internal memory can be the internal memory of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage medium. The readable storage medium can be the computer device's hard drive. In other embodiments, it can also be an external storage device, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, or a flash memory card. Furthermore, memory can include both the computer device's internal storage unit and external storage devices. Memory is used to store the operating system, application programs, boot loaders, data, and other programs, such as the program code of computer programs. Memory can also be used to temporarily store data that has been output or is about to be output.

[0182] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the process steps in the above-mentioned method embodiments by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. Computer-readable media can include at least: any entity or device capable of carrying computer program code, recording media, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signals, telecommunications signals, and software distribution media. Examples include USB flash drives, removable hard drives, magnetic disks, or optical disks. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electric carrier signals or telecommunications signals.

[0183] The present invention may implement all or part of the processes in the above-mentioned method embodiments, and may also be completed through a computer program product. When the computer program product runs on a computer device, the computer device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0184] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0185] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0186] In the embodiments provided by the present invention, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely illustrative. For example, the division of modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0187] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0188] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A method for allocating energy storage resources based on the power Internet of Things, characterized in that: The method comprises: Obtain the historical state of charge sequences corresponding to M energy storage terminals, where M is a positive integer; For any energy storage terminal, a reference state of charge sequence corresponding to the energy storage terminal is obtained based on the historical state of charge sequence corresponding to the energy storage terminal and the trained state of charge prediction model; Determine a dischargeable amount sequence and a chargeable amount sequence corresponding to the energy storage end according to a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, and a maximum amount of charge corresponding to the energy storage end; Determine the overall discharge capacity sequence based on the discharge capacity sequences corresponding to the M energy storage terminals; Determine the overall charge capacity sequence based on the discharge capacity sequences corresponding to the M energy storage terminals; According to the acquired historical load sequence and the trained load forecasting model, the forecast load sequence is obtained; Obtaining a reference load sequence according to the predicted load sequence, the sequence reconstruction model, and the training loss function, wherein obtaining the reference load sequence according to the predicted load sequence, the sequence reconstruction model, and the training loss function includes: Inputting the predicted load sequence into the sequence reconstruction model to obtain a reconstructed load sequence; Calculating a training loss based on the reconstructed load sequence, the predicted load sequence, and the training loss function, wherein calculating a training loss based on the reconstructed load sequence, the predicted load sequence, and the training loss function includes: Accumulating the predicted load sequence to obtain a first load amount; Accumulating the reconstructed load sequence to obtain a second load amount; determining a first sub-loss according to a difference between the first load amount and the second load amount; Using the variance of the reconstructed load sequence as the second sub-loss; Multiplying the first sub-loss by a preset first weight to obtain a first multiplication result; Multiplying the second sub-loss by a preset second weight to obtain a second multiplication result; Taking the sum of the first multiplication result and the second multiplication result as the training loss; Updating the parameters of the sequence reconstruction model according to the training loss until the training loss converges to obtain a trained sequence reconstruction model; Inputting the predicted load sequence into the trained sequence reconstruction model to obtain the reference load sequence; Determine a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determine a reference correction value sequence based on the reference load difference sequence; determining an evaluation parameter of the reference correction amount sequence based on the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence, wherein the reference correction amount sequence includes reference correction amounts corresponding to N preset time points, the overall discharge amount sequence includes overall discharge amounts corresponding to N preset time points, and the overall charge amount sequence includes overall charge amounts corresponding to N preset time points, where N is a positive integer; The step of determining the evaluation parameter of the reference correction amount sequence according to the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence includes: For any preset time point, if the reference correction amount corresponding to the preset time point is less than the overall discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point is greater than the overall charge amount corresponding to the preset time point, determine the sub-evaluation value corresponding to the preset time point as the first preset value; Otherwise, calculating a first difference between the total discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point; Calculating a second difference between the reference correction amount corresponding to the preset time point and the overall charge amount corresponding to the preset time point; Taking the smaller value of the first difference and the second difference as the target value, performing mapping processing on the target value according to the reference correction amount corresponding to the preset time point, and taking the mapping processing result as the sub-evaluation value corresponding to the preset time point; Traversing N preset time points, adding the sub-evaluation values corresponding to the N preset time points, and using the added result as the evaluation parameter of the reference correction value sequence; If the evaluation parameter satisfies a preset condition, determining the reference correction value sequence as the target correction value sequence; If the evaluation parameter does not meet the preset condition, the second weight is updated with a preset step size, and the step of obtaining a reference load sequence based on the predicted load sequence, the sequence reconstruction model and the training loss function is returned to execution; According to the target correction amount sequence, the estimated charge amount or the estimated discharge amount corresponding to each of the M energy storage terminals at a preset time point is determined.

2. The energy storage resource allocation method according to claim 1, characterized in that: The reference state of charge sequence includes reference states of charge corresponding to N preset time points, where N is a positive integer; The step of determining a dischargeable capacity sequence and a chargeable capacity sequence corresponding to the energy storage end according to a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, a state of charge lower limit value corresponding to the energy storage end, and a maximum capacity corresponding to the energy storage end includes: For any preset time point, determine a first state of charge difference based on the state of charge upper limit and the reference state of charge corresponding to the preset time point, and determine the dischargeable capacity of the energy storage end corresponding to the preset time point based on the first state of charge difference and the maximum capacity; Determining a second state-of-charge difference based on the reference state-of-charge corresponding to the preset time point and the lower limit of the state-of-charge, and determining a chargeable capacity of the energy storage terminal corresponding to the preset time point based on the second state-of-charge difference and the maximum capacity; Traversing N preset time points, the dischargeable amounts corresponding to the obtained N preset time points form a dischargeable amount sequence corresponding to the energy storage end, and the chargeable amounts corresponding to the obtained N preset time points form a chargeable amount sequence corresponding to the energy storage end.

3. The energy storage resource allocation method according to claim 1, characterized in that: The target correction amount sequence includes target correction amounts corresponding to N preset time points, the dischargeable amount sequence includes dischargeable amounts corresponding to N preset time points, and the chargeable amount sequence includes chargeable amounts corresponding to N preset time points, where N is a positive integer. The step of determining the estimated charge amounts or discharge amounts corresponding to the M energy storage terminals at preset time points according to the target correction amount sequence includes: For any preset time point, if the target correction amount corresponding to the preset time point is less than the second preset value, then the estimated charge capacity corresponding to each of the M energy storage terminals is determined based on the target correction amount corresponding to the preset time point and the chargeable capacity of each of the M energy storage terminals at the preset time point; If the target correction amount corresponding to the preset time point is greater than or equal to the second preset value, the estimated discharge amount corresponding to the M energy storage ends is determined according to the target correction amount corresponding to the preset time point and the dischargeable amount corresponding to the M energy storage ends at the preset time point.

4. An energy storage resource allocation device based on the power Internet of Things, characterized in that: The device comprises: A state of charge acquisition module is used to obtain the historical state of charge sequences corresponding to M energy storage terminals, where M is a positive integer; The state of charge prediction module is used to obtain a reference state of charge sequence corresponding to any energy storage terminal based on the historical state of charge sequence corresponding to the energy storage terminal and the trained state of charge prediction model; A power determination module is used to determine a dischargeable power sequence and a chargeable power sequence corresponding to the energy storage end based on a reference state of charge sequence corresponding to the energy storage end, a state of charge upper limit value corresponding to the energy storage end, a state of charge lower limit value corresponding to the energy storage end, and a maximum power corresponding to the energy storage end; A first sequence determination module is used to determine an overall discharge capacity sequence based on the dischargeable capacity sequences corresponding to the M energy storage terminals; A second sequence determination module is used to determine the overall charge capacity sequence according to the dischargeable capacity sequences corresponding to the M energy storage terminals; The load forecasting module is used to obtain the predicted load sequence based on the acquired historical load sequence and the trained load forecasting model; A sequence reconstruction module is configured to obtain a reference load sequence based on the predicted load sequence, the sequence reconstruction model, and the training loss function, wherein obtaining the reference load sequence based on the predicted load sequence, the sequence reconstruction model, and the training loss function includes: Inputting the predicted load sequence into the sequence reconstruction model to obtain a reconstructed load sequence; Calculating a training loss based on the reconstructed load sequence, the predicted load sequence, and the training loss function. Calculating a training loss based on the reconstructed load sequence, the predicted load sequence, and the training loss function includes: Accumulating the predicted load sequence to obtain a first load amount; Accumulating the reconstructed load sequence to obtain a second load amount; determining a first sub-loss according to a difference between the first load amount and the second load amount; Using the variance of the reconstructed load sequence as the second sub-loss; Multiplying the first sub-loss by a preset first weight to obtain a first multiplication result; Multiplying the second sub-loss by a preset second weight to obtain a second multiplication result; Taking the sum of the first multiplication result and the second multiplication result as the training loss; Updating the parameters of the sequence reconstruction model according to the training loss until the training loss converges to obtain a trained sequence reconstruction model; Inputting the predicted load sequence into the trained sequence reconstruction model to obtain the reference load sequence; A correction value determination module is used to determine a reference load difference sequence based on the predicted load sequence and the reference load sequence, and determine a reference correction value sequence based on the reference load difference sequence; a sequence evaluation module, configured to determine an evaluation parameter of the reference correction amount sequence based on the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence, wherein the reference correction amount sequence includes reference correction amounts corresponding to N preset time points, the overall discharge amount sequence includes overall discharge amounts corresponding to N preset time points, and the overall charge amount sequence includes overall charge amounts corresponding to N preset time points, where N is a positive integer; The step of determining the evaluation parameter of the reference correction amount sequence according to the reference correction amount sequence, the overall discharge amount sequence, and the overall charge amount sequence includes: For any preset time point, if the reference correction amount corresponding to the preset time point is less than the overall discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point is greater than the overall charge amount corresponding to the preset time point, determine the sub-evaluation value corresponding to the preset time point as the first preset value; Otherwise, calculating a first difference between the total discharge amount corresponding to the preset time point and the reference correction amount corresponding to the preset time point; Calculating a second difference between the reference correction amount corresponding to the preset time point and the overall charge amount corresponding to the preset time point; Taking the smaller value of the first difference and the second difference as the target value, performing mapping processing on the target value according to the reference correction amount corresponding to the preset time point, and taking the mapping processing result as the sub-evaluation value corresponding to the preset time point; Traversing N preset time points, adding the sub-evaluation values corresponding to the N preset time points, and using the added result as the evaluation parameter of the reference correction value sequence; a first condition judgment module, configured to determine the reference correction value sequence as a target correction value sequence if the evaluation parameter satisfies a preset condition; A second condition judgment module is configured to update the second weight with a preset step size if the evaluation parameter does not meet the preset condition, and return to the step of obtaining a reference load sequence based on the predicted load sequence, the sequence reconstruction model, and the training loss function; The power distribution module is used to determine the estimated charge capacity or estimated discharge capacity corresponding to M energy storage terminals at preset time points according to the target correction amount sequence.

5. A computer device, characterized in that: The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the energy storage resource allocation method based on the power Internet of Things as described in any one of claims 1 to 3 is implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the energy storage resource allocation method based on the power Internet of Things as described in any one of claims 1 to 3 is implemented.

Images