Compressed air energy storage control method and system

Abstract

The present invention relates to the field of air energy storage technology and discloses a compressed air energy storage control method and system. Within a first preset time period, power consumption information of a target power storage system is collected at fixed time intervals; the energy storage tanks of the target power storage system are divided, specifically including: a first division for using M energy storage tanks as buffer units; a second division for using N energy storage tanks as energy storage units; wherein the buffer unit is used to store the input power of the target power storage system at a first air compression ratio; and the energy storage unit is used to store the input power of the target power storage system at a second air compression ratio; power consumption information at K moments within the first preset time period is processed and analyzed to obtain predicted input power and predicted output power of the target power storage system in a second preset time period; and dynamic energy storage planning is performed on the buffer unit and the energy storage unit based on the predicted input power and predicted output power of the target power storage system in the second preset time period.

Description

Technical Field

[0001] The present invention relates to the technical field of air energy storage, and more particularly, to a compressed air energy storage control method and system. Background Art

[0002] An air energy storage system compresses air to a high pressure using a compressor and stores it in a tank. Air energy storage systems are characterized by large capacity, long life, and environmental friendliness, and are widely used in supporting energy storage for renewable energy generation. However, air energy storage systems still have the following problems:

[0003] 1. During the operation of the air energy storage system, due to the instability of power input or changes in power load, the compressor needs to be started and stopped frequently, which makes it difficult for the compressor to respond to power demand in a timely manner. At the same time, frequent starting and stopping can easily lead to mechanical failures.

[0004] 2. When the air energy storage system is in operation, in order to improve the energy storage efficiency, it is necessary to give priority to energy storage tanks with lower temperatures for compressed air energy storage. However, the temperatures of the energy storage tanks are not consistent, and the heat emitted by the energy storage tanks will interfere with each other, making it difficult to select the energy storage tanks. Summary of the Invention

[0005] The present invention provides a compressed air energy storage control method and system to solve the technical problems raised in the background technology.

[0006] The present invention provides a compressed air energy storage control system, comprising:

[0007] a data collection module configured to collect electricity usage information of the target electricity storage system at fixed time intervals during a first preset time period; wherein the electricity usage information includes: input power, output power, temperature and weather of the region to which the input power terminal of the target electricity storage system belongs, and temperature and weather of the region to which the output power terminal of the target electricity storage system belongs;

[0008] The energy storage tank division module is used to divide the energy storage tanks of the target power storage system, specifically including: a first division for dividing M energy storage tanks as buffer units; a second division for dividing N energy storage tanks as energy storage units;

[0009] The buffer unit is used to store the input power of the target power storage system at a first air compression ratio; the energy storage unit is used to store the input power of the target power storage system at a second air compression ratio; and the buffer unit is used to keep the target power storage system running at no load at any time during a preset time period.

[0010] a data analysis module, configured to process and analyze the electricity consumption information at K moments in a first preset time period to obtain a predicted input power and a predicted output power of the target electricity storage system in a second preset time period;

[0011] The dynamic planning module is used to perform dynamic energy storage planning for the buffer unit and the energy storage unit based on the predicted input power and predicted output power of the target power storage system in the second preset time period.

[0012] Furthermore, the energy storage tanks of the target electricity storage system are divided into:

[0013] Divide the daytime period into the nighttime period based on manual experience, use the daytime period as the preset period, and obtain the compressed no-load power consumption of the energy storage unit of the target power storage system during the preset period;

[0014] Based on the compressed no-load power consumption of the energy storage unit during the daytime, the buffer unit and energy storage unit are divided. The division formula is as follows:

[0015] ;

[0016] in, Indicates the number of energy storage tanks in the buffer unit, Indicates the number of energy storage tanks in the energy storage unit, Indicates the ratio of the energy storage tank's compressed no-load power consumption to its compressed full-load power consumption. Indicates the compression efficiency of the energy storage tank , Indicates the adiabatic index of air , represents the first air compression ratio, represents the second air compression ratio, Indicates the length of the daytime period.

[0017] Furthermore, the power consumption information at K moments in the first preset time period is processed and analyzed, including:

[0018] The electricity consumption information at K moments is preprocessed. The preprocessing includes:

[0019] For missing values ​​in the electricity consumption information, the average value of the electricity consumption information of the two adjacent time points corresponding to the missing value is used to fill the missing value, and the electricity consumption information is normalized to obtain the standard electricity consumption information of K time points;

[0020] Based on the standard electricity consumption information at each moment, an electricity consumption feature vector is constructed. A fixed-length electricity consumption feature vector is extracted through a sliding window to construct an electricity consumption feature matrix as a training sample. The input and output power at the next moment of the sliding window are used as the sample labels of the corresponding training samples.

[0021] Based on the training samples and the sample labels, a power prediction model is trained; the power prediction model includes a hidden layer, a first classifier, and a second classifier;

[0022] The hidden layer is used to extract the hidden state of the power consumption feature matrix. The classification space of the first classifier represents the output power of the corresponding power consumption feature matrix, and the second classification space represents the input power of the corresponding power consumption feature matrix.

[0023] Furthermore, the hidden layer of the power consumption prediction model includes hidden units. The number of hidden units is the same as the number of columns in the power consumption feature matrix. Each hidden unit is used to extract the independent state of a column of data in the power consumption feature matrix. The hidden units are all constructed based on a 1D convolutional neural network.

[0024] The calculation formula of the hidden layer is as follows:

[0025] ;

[0026] in, represents the hidden state of the hidden layer output, represents the average pooling operation, express activation function, represents the input electricity consumption feature matrix, represents the number of columns of the electricity consumption feature matrix, express The index of Represents the electricity consumption characteristic matrix The convolution kernel of the column, Represents a convolution operation.

[0027] Furthermore, dynamic energy storage planning is performed on the buffer unit and the energy storage unit, including:

[0028] If the difference S between the predicted input power and the predicted output power in the second preset time period is greater than K; it is determined that all M energy storage tanks in the buffer unit have reached the first air compression ratio, then the first energy storage scheme is activated, otherwise the second energy storage scheme is activated;

[0029] If the difference S between the predicted input power and the predicted output power in the second preset time period is less than or equal to K, the third energy storage solution is activated;

[0030] Wherein, K>0, K represents the power required for the N energy storage tanks in the energy storage unit to maintain compressed no-load power consumption within the second preset time period.

[0031] Furthermore, the first energy storage solution includes inputting the power difference S into the N energy storage tanks of the energy storage unit:

[0032] Constructing an energy storage network diagram for the energy storage unit. The energy storage network diagram includes: mapping N energy storage tanks in the energy storage unit of the target electric storage system into energy storage nodes; if any two energy storage tanks in the energy storage unit are adjacent, establishing a topological edge between the energy storage nodes corresponding to the adjacent energy storage tanks;

[0033] The temperature and actual air compression ratio of each energy storage tank in the energy storage unit within the second preset time period are obtained, and characteristic parameters of the corresponding energy storage node are calculated based on the temperature and actual air compression ratio of the energy storage tank. The calculation formula of the characteristic parameters is as follows:

[0034] ;

[0035] in, represents the characteristic parameters of the j-th energy storage node, represents the second air compression ratio, represents the actual air compression ratio of the jth energy storage tank in the energy storage unit, represents the temperature of the jth energy storage tank in the energy storage unit;

[0036] Sort the characteristic parameters of each energy storage node from small to large to obtain a characteristic ranking;

[0037] The energy storage tank corresponding to each energy storage node in the feature sorting is compressed and unloaded within a second preset time period; a target energy storage tank is set and the remaining power SK is stored;

[0038] The energy storage node with the first rank in the feature sorting is selected as the target energy storage tank in the second preset time period, and the energy storage network diagram is updated in the third preset time period. The update formula of the energy storage network diagram is as follows:

[0039] ;

[0040] ;

[0041] in, represents the updated characteristic parameters of the j-th energy storage node, represents the updated air compression ratio of the j-th energy storage tank in the energy storage unit, Parameters indicating the effect of temperature on compression ratio, Indicates the power of air compression charging, represents the temperature rise of the j-th energy storage tank, represents the heat capacity of the j-th energy storage tank, represents the temperature of the j-th energy storage tank, represents the updated characteristic parameters of the i-th energy storage node, represents the actual air compression ratio of the i-th energy storage tank in the energy storage unit, represents the temperature of the i-th energy storage tank, represents the heat diffusion transfer coefficient, , e represents the natural base, represents the minimum number of hops between the i-th energy storage node and the j-th energy storage node in the energy storage network graph, represents the hop weight parameter, represents the Euclidean distance between the i-th energy storage tank and the j-th energy storage tank, represents the updated temperature of the j-th energy storage tank, Indicates the ambient temperature.

[0042] Furthermore, the second energy storage solution includes taking the energy storage tank in the buffer unit that has not reached the first air compression ratio as the target energy storage tank during the second preset time period.

[0043] Furthermore, the third energy storage solution includes expanding the M energy storage tanks in the buffer unit to obtain backup power; and distributing the backup power to the N energy storage tanks in the energy storage unit to keep the N energy storage tanks compressed and unloaded.

[0044] In a second aspect, a compressed air energy storage control method is applied to the system, comprising:

[0045] Step 1: Collecting electricity usage information of a target electricity storage system at fixed time intervals within a first preset time period;

[0046] Step 2: Divide the energy storage tanks of the target electricity storage system;

[0047] Step 3: Process and analyze the power consumption information at K moments in the first preset time period to obtain the predicted input power and predicted output power of the target power storage system in the second preset time period;

[0048] Step 4: Based on the predicted input power and predicted output power of the target power storage system in the second preset time period, dynamic energy storage planning is performed on the buffer unit and the energy storage unit.

[0049] The beneficial effects of this invention are that by dividing the target power storage system into buffer units and energy storage units, and combining power forecasting models with dynamic energy storage planning strategies, it not only effectively reduces the frequent start-stop and stop of compressors in the target power storage system, extending equipment life, but also improves energy storage efficiency and system operational stability. Furthermore, by introducing neural network-based power consumption information analysis, power demand forecasting is achieved, enabling the energy storage system to flexibly adapt to fluctuating power storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a module diagram of a compressed air energy storage control system of the present invention;

[0051] Figure 2 It is a flow chart of a compressed air energy storage control method of the present invention. DETAILED DESCRIPTION

[0052] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed solely to enable those skilled in the art to better understand and implement the subject matter described herein, and that the functions and arrangements of the elements discussed may be varied without departing from the scope of this specification. Various examples may omit, substitute, or add various processes or components as needed. In addition, features described with respect to some examples may also be combined in other examples.

[0053] like Figure 1 As shown, a compressed air energy storage control system includes:

[0054] a data collection module configured to collect electricity usage information of the target electricity storage system at fixed time intervals during a first preset time period; wherein the electricity usage information includes: input power, output power, temperature and weather of the region to which the input power terminal of the target electricity storage system belongs, and temperature and weather of the region to which the output power terminal of the target electricity storage system belongs;

[0055] The energy storage tank division module is used to divide the energy storage tanks of the target power storage system, specifically including: a first division for dividing M energy storage tanks as buffer units; a second division for dividing N energy storage tanks as energy storage units;

[0056] The buffer unit is used to store the input power of the target power storage system at a first air compression ratio; the energy storage unit is used to store the input power of the target power storage system at a second air compression ratio; and the buffer unit is used to keep the target power storage system running at no load at any time during a preset time period.

[0057] a data analysis module, configured to process and analyze the electricity consumption information at K moments in a first preset time period to obtain a predicted input power and a predicted output power of the target electricity storage system in a second preset time period;

[0058] The dynamic planning module is used to perform dynamic energy storage planning for the buffer unit and the energy storage unit based on the predicted input power and predicted output power of the target power storage system in the second preset time period.

[0059] Compressed air energy storage (CAES) does not impose special requirements on the material of the energy storage equipment used in energy storage scenarios, so compressed air can be stored in artificial steel tanks. This feature makes CAES technology very suitable for use in conjunction with photovoltaic power generation systems in industrial parks or factories. It stores electricity generated by photovoltaic power generation during the day and releases the stored electricity for use at night. However, due to the instability of photovoltaic power generation, the compressor in the CAES system may start and stop frequently. Frequent starting and stopping of the compressor not only significantly increases the energy consumption when restarting, but also significantly shortens its service life due to the greater force exerted on the compressor under high compression ratio conditions. In addition, the operating characteristics of the energy storage tank are different under high and low compression ratio conditions, so the system needs to be rationally optimized.

[0060] To address the above issues, this solution proposes a strategy for the layered use of energy storage tanks: the energy storage tanks are divided into buffer units and energy storage units. The energy storage tanks of the buffer units store a small amount of electricity at a lower air compression ratio, which is mainly used to smooth out power input fluctuations and keep the energy storage units running at no load in sunlight-free conditions (such as cloudy days or at night); the energy storage tanks of the energy storage units store the main amount of electricity at a higher air compression ratio, ensuring the efficient and stable operation of the system under high compression ratio conditions. Since the buffer unit undertakes the main start-stop operations, its lower compression ratio effectively reduces the mechanical burden and failure risk of the compressor, while avoiding direct impact on the operation of the energy storage unit. Therefore, this strategy not only improves the overall stability and response speed of the system, but also extends the service life of the equipment and optimizes the energy storage efficiency.

[0061] In one embodiment of the present invention, dividing the energy storage tanks of the target electricity storage system includes:

[0062] Divide the daytime period into the nighttime period based on manual experience, use the daytime period as the preset period, and obtain the compressed no-load power consumption of the energy storage unit of the target power storage system during the preset period;

[0063] Based on the compressed no-load power consumption of the energy storage unit during the daytime, the buffer unit and energy storage unit are divided. The division formula is as follows:

[0064] ;

[0065] in, Indicates the number of energy storage tanks in the buffer unit, Indicates the number of energy storage tanks in the energy storage unit, Indicates the ratio of the energy storage tank's compressed no-load power consumption to its compressed full-load power consumption. Indicates the compression efficiency of the energy storage tank , Indicates the adiabatic index of air , represents the first air compression ratio, represents the second air compression ratio, Indicates the length of the daytime period.

[0066] Specifically, a partitioning formula divides the number of energy storage tanks in any scenario, thereby stably selecting a portion of energy storage tanks as buffer units and another portion as energy storage units. The buffer units also ensure that the energy storage units can operate at no load at any time during the day, enhancing the stability of the energy storage system.

[0067] In one embodiment of the present invention, processing and analyzing the electricity usage information at K moments in a first preset time period includes:

[0068] The electricity consumption information at K moments is preprocessed. The preprocessing includes:

[0069] For missing values ​​in the electricity consumption information, the average value of the electricity consumption information of the two adjacent time points corresponding to the missing value is used to fill the missing value, and the electricity consumption information is normalized to obtain the standard electricity consumption information of K time points;

[0070] Based on the standard electricity consumption information at each moment, an electricity consumption feature vector is constructed. A fixed-length electricity consumption feature vector is extracted through a sliding window to construct an electricity consumption feature matrix as a training sample. The input and output power at the next moment of the sliding window are used as the sample labels of the corresponding training samples.

[0071] Based on the training samples and the sample labels, a power prediction model is trained; the power prediction model includes a hidden layer, a first classifier, and a second classifier;

[0072] The hidden layer is used to extract the hidden state of the power consumption feature matrix. The classification space of the first classifier represents the output power of the corresponding power consumption feature matrix, and the second classification space represents the input power of the corresponding power consumption feature matrix.

[0073] Specifically, the power forecasting model can accurately predict the input power and output power of the next time period. For example, a set of data includes , the size of the sliding window is 3, then The corresponding electricity consumption feature matrix is ​​used as a training sample. The input and output power in the data are used as the corresponding sample labels. Several sets of training samples and sample labels are obtained through a loop to train a power prediction model. The power prediction model uses backpropagation to update its hidden layer hyperparameters using a mean squared error loss function to achieve accurate predictions.

[0074] It should be noted that the length of the first preset time period is much longer than the second preset time period and the subsequent third preset time period.

[0075] In one embodiment of the present invention, the hidden layer of the power prediction model includes hidden units. The number of hidden units is the same as the number of columns in the power consumption feature matrix. Each hidden unit is used to extract the independent state of a column of data in the power consumption feature matrix. The hidden units are all constructed based on a 1D convolutional neural network.

[0076] The calculation formula of the hidden layer is as follows:

[0077] ;

[0078] in, represents the hidden state of the hidden layer output, represents the average pooling operation, express activation function, represents the input electricity consumption feature matrix, represents the number of columns of the electricity consumption feature matrix, express The index of Represents the electricity consumption characteristic matrix The convolution kernel of the column, Represents a convolution operation.

[0079] Specifically, the hidden layer includes G hidden units, each corresponding to a single type of data. Each hidden unit is assigned a one-dimensional convolution kernel and stride. This allows convolution pooling of each column of the power feature matrix to obtain its independent state. The independent states corresponding to each column are concatenated to obtain the hidden state of the power feature matrix. These hidden states are input into the first and second classifiers to obtain the predicted output and input power, respectively.

[0080] In one embodiment of the present invention, dynamic energy storage planning is performed on the buffer unit and the energy storage unit, including:

[0081] If the difference S between the predicted input power and the predicted output power in the second preset time period is greater than K; it is determined that all M energy storage tanks in the buffer unit have reached the first air compression ratio, then the first energy storage scheme is activated, otherwise the second energy storage scheme is activated;

[0082] If the difference S between the predicted input power and the predicted output power in the second preset time period is less than or equal to K, the third energy storage solution is activated;

[0083] Wherein, K>0, K represents the power required for the N energy storage tanks in the energy storage unit to maintain compressed no-load power consumption within the second preset time period.

[0084] Specifically, there are three options for energy storage in the energy storage unit. All three options prioritize replenishing the buffer unit to ensure that the energy storage tank in the energy storage unit can operate at no load at any time.

[0085] In one embodiment of the present invention, the first energy storage solution includes inputting the power difference S into N energy storage tanks of the energy storage unit:

[0086] Constructing an energy storage network diagram for the energy storage unit. The energy storage network diagram includes: mapping N energy storage tanks in the energy storage unit of the target electric storage system into energy storage nodes; if any two energy storage tanks in the energy storage unit are adjacent, establishing a topological edge between the energy storage nodes corresponding to the adjacent energy storage tanks;

[0087] The temperature and actual air compression ratio of each energy storage tank in the energy storage unit within the second preset time period are obtained, and characteristic parameters of the corresponding energy storage node are calculated based on the temperature and actual air compression ratio of the energy storage tank. The calculation formula of the characteristic parameters is as follows:

[0088] ;

[0089] in, represents the characteristic parameters of the j-th energy storage node, represents the second air compression ratio, represents the actual air compression ratio of the jth energy storage tank in the energy storage unit, represents the temperature of the jth energy storage tank in the energy storage unit;

[0090] Sort the characteristic parameters of each energy storage node from small to large to obtain a characteristic ranking;

[0091] The energy storage tank corresponding to each energy storage node in the feature sorting is compressed and unloaded within a second preset time period; a target energy storage tank is set and the remaining power SK is stored;

[0092] The energy storage node with the first rank in the feature sorting is selected as the target energy storage tank in the second preset time period, and the energy storage network diagram is updated in the third preset time period. The update formula of the energy storage network diagram is as follows:

[0093] ;

[0094] ;

[0095] in, represents the updated characteristic parameters of the j-th energy storage node, represents the updated air compression ratio of the j-th energy storage tank in the energy storage unit, Parameters indicating the effect of temperature on compression ratio, Indicates the power of air compression charging, represents the temperature rise of the j-th energy storage tank, represents the heat capacity of the j-th energy storage tank, represents the temperature of the j-th energy storage tank, represents the updated characteristic parameters of the i-th energy storage node, represents the actual air compression ratio of the i-th energy storage tank in the energy storage unit, represents the temperature of the i-th energy storage tank, represents the heat diffusion transfer coefficient, , e represents the natural base, represents the minimum number of hops between the i-th energy storage node and the j-th energy storage node in the energy storage network graph, represents the hop weight parameter, represents the Euclidean distance between the i-th energy storage tank and the j-th energy storage tank, represents the updated temperature of the j-th energy storage tank, Indicates the ambient temperature.

[0096] In one embodiment of the present invention, The value range is ; The value range is ; The value range is .

[0097] In one embodiment of the present invention, the formulas used in this application are calculated based on dimensionless values. Dimensional values ​​are normalized based on standard values. For example, the standard value for quantity is ten, the standard value for daytime duration is one hour, the standard value for temperature is five degrees Celsius, and the standard value for power is one kilowatt. After obtaining the corresponding parameters, the ratio of the parameters to the standard values ​​is calculated and normalized. The normalized parameters are then dimensionally removed for calculation.

[0098] Specifically, the characteristic parameters of the energy storage node corresponding to each energy storage tank are calculated by obtaining the temperature and actual air compression ratio of each energy storage tank in the energy storage unit. These characteristic parameters are used for sorting and selecting the most suitable energy storage tank as the target tank. The greater the difference between the air compression ratio and the expected compression ratio, the more suitable the target tank is, and the lower the temperature, the more suitable the target tank is.

[0099] Specifically, after the target energy storage tank completes energy storage during the second preset time period, the gas in the target tank is compressed and heated, and the air compression ratio increases. After the third preset time period, the characteristic parameters of the target energy storage tank need to be updated. At the same time, due to the arrangement of the energy storage tanks, the target energy storage tank will continuously dissipate heat, causing the surrounding energy storage tanks to be continuously affected by the temperature. To mitigate this effect, compensation is required based on the distance between the target energy storage tank and the number of tanks in between (minimum hop count). This allows the characteristic parameters of the other energy storage tanks to be updated, resulting in an updated energy storage network diagram. During the third preset time period, the tanks are sorted again based on the updated energy storage network diagram, and the search for the target energy storage tank continues in this cycle. This ensures that the target energy storage tank is the optimal choice within each time period.

[0100] In one embodiment of the present invention, the second energy storage solution includes selecting an energy storage tank in the buffer unit that has not reached the first air compression ratio as a target energy storage tank during the second preset time period.

[0101] In one embodiment of the present invention, the third energy storage solution includes expanding M energy storage tanks in the buffer unit to obtain backup power; and distributing the backup power to N energy storage tanks in the energy storage unit to keep the N energy storage tanks compressed and unloaded.

[0102] It should be noted that "compression no-load" refers to the compressor operating at no load and can be considered a pre-run operation, without any actual air compression effect. No-load operation consumes approximately 10% to 20% of the actual compression power consumption. No-load operation helps avoid frequent compression starts and stops, extending the life of the compressor associated with the energy storage tank in the energy storage unit.

[0103] In one embodiment of the present invention, the formula of the present invention is calculated based on dimensionless.

[0104] A compressed air energy storage control method, applied to the above system, comprising:

[0105] Step 1: Collecting electricity usage information of a target electricity storage system at fixed time intervals within a first preset time period;

[0106] Step 2: Divide the energy storage tanks of the target electricity storage system;

[0107] Step 3: Process and analyze the power consumption information at K moments in the first preset time period to obtain the predicted input power and predicted output power of the target power storage system in the second preset time period;

[0108] Step 4: Based on the predicted input power and predicted output power of the target power storage system in the second preset time period, dynamic energy storage planning is performed on the buffer unit and the energy storage unit.

[0109] It should be noted that the intervals and thresholds are set for ease of comparison. The threshold size depends on the amount of sample data and the cardinality set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless numerical calculations. These formulas are derived from software simulations of the most recent real-world conditions using large amounts of data. The preset parameters in these formulas are set by those skilled in the art based on actual conditions.

[0110] The above describes the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation methods. The above specific implementation methods are merely illustrative and not restrictive. Ordinary technicians in this field can also make many forms based on the inspiration of this embodiment, all of which are protected by this embodiment.

Claims

1. A compressed air energy storage control system, characterized in that: include: a data collection module configured to collect electricity usage information of the target electricity storage system at fixed time intervals during a first preset time period; wherein the electricity usage information includes: input power, output power, temperature and weather of the region to which the input power terminal of the target electricity storage system belongs, and temperature and weather of the region to which the output power terminal of the target electricity storage system belongs; The energy storage tank division module is used to divide the energy storage tanks of the target power storage system, specifically including: a first division for dividing M energy storage tanks as buffer units; a second division for dividing N energy storage tanks as energy storage units; The buffer unit is used to store the input power of the target power storage system at a first air compression ratio; the energy storage unit is used to store the input power of the target power storage system at a second air compression ratio; and the buffer unit is used to keep the target power storage system running at no load at any time during a preset time period. a data analysis module, configured to process and analyze the electricity consumption information at K moments in a first preset time period to obtain a predicted input power and a predicted output power of the target electricity storage system in a second preset time period; A dynamic planning module is used to perform dynamic energy storage planning for the buffer unit and the energy storage unit based on the predicted input power and predicted output power of the target energy storage system in the second preset time period, including: If the difference S between the predicted input power and the predicted output power in the second preset time period is greater than H; H>0, H represents the power required for the N energy storage tanks in the energy storage unit to maintain compressed no-load power consumption in the second preset time period; if it is determined that all M energy storage tanks in the buffer unit have reached the first air compression ratio, then the first energy storage plan is activated, otherwise the second energy storage plan is activated; if the difference S between the predicted input power and the predicted output power in the second preset time period is less than or equal to H, then the third energy storage plan is activated; The first energy storage solution includes inputting the power difference S into N energy storage tanks of the energy storage unit, including: Constructing an energy storage network diagram for the energy storage unit. The energy storage network diagram includes: mapping N energy storage tanks in the energy storage unit of the target electric storage system into energy storage nodes; if any two energy storage tanks in the energy storage unit are adjacent, establishing a topological edge between the energy storage nodes corresponding to the adjacent energy storage tanks; Obtaining the temperature and actual air compression ratio of each energy storage tank in the energy storage unit within a second preset time period, and calculating characteristic parameters of the corresponding energy storage node based on the temperature and actual air compression ratio of the energy storage tank; Sort the characteristic parameters of each energy storage node from small to large to obtain a characteristic ranking; The energy storage tank corresponding to each energy storage node in the feature sorting is compressed and unloaded within a second preset time period; a target energy storage tank is set to store the remaining power SH; The energy storage node with the first rank in the feature sorting is selected as the target energy storage tank in the second preset time period, and the energy storage network diagram is updated in the third preset time period.

2. A compressed air energy storage control system according to claim 1, characterized in that: The energy storage tanks of the target electricity storage system are divided into: Divide the daytime period into the nighttime period based on manual experience, use the daytime period as the preset period, and obtain the compressed no-load power consumption of the energy storage unit of the target power storage system during the preset period; Based on the compressed no-load power consumption of the energy storage unit during the daytime, the buffer unit and energy storage unit are divided. The division formula is as follows: ； in, Indicates the number of energy storage tanks in the buffer unit, Indicates the number of energy storage tanks in the energy storage unit, Indicates the ratio of the energy storage tank's compressed no-load power consumption to its compressed full-load power consumption. Indicates the compression efficiency of the energy storage tank , Indicates the adiabatic index of air , represents the first air compression ratio, represents the second air compression ratio, Indicates the length of the daytime period.

3. A compressed air energy storage control system according to claim 1, characterized in that: The power consumption information at K moments in the first preset time period is processed and analyzed, including: The electricity consumption information at K moments is preprocessed. The preprocessing includes: For missing values ​​in the electricity consumption information, the average value of the electricity consumption information of the two adjacent time points corresponding to the missing value is used to fill the missing value, and the electricity consumption information is normalized to obtain the standard electricity consumption information of K time points; Based on the standard electricity consumption information at each moment, an electricity consumption feature vector is constructed. A fixed-length electricity consumption feature vector is extracted through a sliding window to construct an electricity consumption feature matrix as a training sample. The input and output power at the next moment of the sliding window are used as the sample labels of the corresponding training samples. Based on the training samples and the sample labels, a power prediction model is trained; the power prediction model includes a hidden layer, a first classifier, and a second classifier; The hidden layer is used to extract the hidden state of the power consumption feature matrix. The classification space of the first classifier represents the output power of the corresponding power consumption feature matrix, and the second classification space represents the input power of the corresponding power consumption feature matrix.

4. A compressed air energy storage control system according to claim 3, characterized in that: The hidden layer of the power consumption prediction model includes hidden units. The number of hidden units is the same as the number of columns in the power consumption feature matrix. Each hidden unit is used to extract the independent state of a column of data in the power consumption feature matrix. The hidden units are all constructed based on a 1D convolutional neural network. The calculation formula of the hidden layer is as follows: ； in, represents the hidden state of the hidden layer output, represents the average pooling operation, express activation function, represents the input electricity consumption feature matrix, represents the number of columns of the electricity consumption feature matrix, express The index of Represents the electricity consumption characteristic matrix The convolution kernel of the column, Represents a convolution operation.

5. A compressed air energy storage control system according to claim 1, characterized in that: The calculation formula of characteristic parameters is as follows: ； in, represents the characteristic parameter of the j-th energy storage node, represents the second air compression ratio, represents the actual air compression ratio of the jth energy storage tank in the energy storage unit, Represents the temperature of the j-th energy storage tank in the energy storage unit.

6. A compressed air energy storage control system according to claim 1, characterized in that: The update formula of the energy storage network graph is as follows: ； ； in, represents the updated characteristic parameters of the j-th energy storage node, represents the updated air compression ratio of the j-th energy storage tank in the energy storage unit, Parameters indicating the effect of temperature on compression ratio, Indicates the power of air compression charging, represents the temperature rise of the j-th energy storage tank, represents the heat capacity of the j-th energy storage tank, represents the temperature of the j-th energy storage tank, represents the updated characteristic parameters of the i-th energy storage node, represents the actual air compression ratio of the i-th energy storage tank in the energy storage unit, represents the temperature of the i-th energy storage tank, represents the heat diffusion transfer coefficient, , e represents the natural base, represents the minimum number of hops between the i-th energy storage node and the j-th energy storage node in the energy storage network graph, represents the hop weight parameter, represents the Euclidean distance between the i-th energy storage tank and the j-th energy storage tank, represents the updated temperature of the j-th energy storage tank, Indicates the ambient temperature.

7. A compressed air energy storage control system according to claim 1, characterized in that: The second energy storage solution includes taking the energy storage tank in the buffer unit that has not reached the first air compression ratio as the target energy storage tank in the second preset time period.

8. A compressed air energy storage control system according to claim 1, characterized in that: The third energy storage solution includes expanding the M energy storage tanks in the buffer unit to obtain backup power; and distributing the backup power to the N energy storage tanks in the energy storage unit to keep the N energy storage tanks compressed and unloaded.

9. A compressed air energy storage control method, applied to the system according to any one of claims 1 to 8, characterized in that: include: Step 1: Collecting electricity usage information of a target electricity storage system at fixed time intervals within a first preset time period; Step 2: Divide the energy storage tanks of the target electricity storage system; Step 3: Process and analyze the power consumption information at K moments in the first preset time period to obtain the predicted input power and predicted output power of the target power storage system in the second preset time period; Step 4: Based on the predicted input power and predicted output power of the target power storage system in the second preset time period, dynamic energy storage planning is performed on the buffer unit and the energy storage unit.

Images