Distributed new energy storage system

Abstract

The application discloses a distributed new energy storage system, and belongs to the technical field of new energy storage control, which comprises a new energy power generation sensing module, which is installed on a new energy power generation device, performs new energy power generation sensing, and obtains an output power fluctuation prediction trend in a prediction period; a multi-dimensional state monitoring module, which is installed on a storage unit, and collects the SOC, SOH and cell temperature of the storage unit in real time; a regional coordination node, which receives data collected by the new energy power generation sensing module and the multi-dimensional state monitoring module, outputs a charging call instruction according to the output power fluctuation prediction trend and the SOC, SOH and cell temperature, and controls the power of energy transmitted by the new energy power generation device to different storage units, so that the accuracy of the output power fluctuation prediction trend can be improved, the storage units can be balanced in energy storage work, only a small number of storage units can be made to bear peak loads for a long time, the storage performance can be optimized, and the service life can be prolonged.

Description

Technical Field

[0001] This application belongs to the field of new energy storage control technology, specifically, it relates to a distributed new energy storage system. Background Technology

[0002] With the transformation of the global energy structure and the rapid development of new energy technologies, distributed new energy storage systems have become one of the key technologies for achieving sustainable energy development. Distributed new energy has received widespread attention due to its clean and renewable characteristics. However, new energy power generation is significantly affected by the natural environment, and its output power has strong fluctuations and intermittency. These fluctuations and intermittency of energy pose challenges to the stable operation of the power grid and the effective utilization of energy. At the same time, with the expansion of new energy installed capacity, a single energy storage unit can hardly meet the energy regulation needs in complex scenarios.

[0003] Although existing technologies have made some progress in the management and scheduling of distributed renewable energy storage systems, there are still many shortcomings. Traditional forecasting methods rely heavily on single environmental parameters or simple trend analysis, lacking sufficient flexibility and intelligence. They also lack effective processing of abnormal data and dynamic correction of forecast results, leading to large forecast deviations and affecting the accuracy of energy storage scheduling. The energy allocation strategy of energy storage units is not refined enough, failing to combine the predicted trend of renewable energy power fluctuations with the multi-dimensional status of energy storage units for dynamic weight allocation. This results in an imperfect collaborative working mechanism for energy storage units, leading to low utilization of some units, prolonged use of other units, long-term peak load bearing, shortened lifespan, and potentially different lifespans among units installed in the same batch, making unified maintenance and replacement difficult and increasing work inefficiency. Moreover, the collaborative control capability between energy storage units is weak, lacking a global power allocation benchmark and dynamic balancing mechanism. When the status differences between energy storage units are large, it is difficult to achieve efficient energy flow and complementarity, affecting the overall stability of the system. Summary of the Invention

[0004] To address the aforementioned problems and technical deficiencies, this application adopts the following technical solution: a distributed new energy storage system, comprising:

[0005] The new energy power generation sensing module is installed on new energy power generation equipment to sense new energy power generation and obtain the output power fluctuation prediction trend within the prediction period.

[0006] A multi-dimensional status monitoring module is installed on the energy storage unit to collect the SOC, SOH and cell temperature of the energy storage unit in real time;

[0007] The regional coordination node is used to receive data collected by the new energy power generation sensing module and the multi-dimensional status monitoring module, and output charging call commands based on the output power fluctuation prediction trend and SOC, SOH and cell temperature to control the power of the new energy power generation equipment to transmit energy to different energy storage units.

[0008] Preferably, the new energy power generation sensing includes:

[0009] Photosensitive sensors and temperature sensors are installed on photovoltaic power generation equipment, and wind speed sensors are installed on wind power generation equipment. The sensors are used to collect environmental data of the new energy power generation equipment during the detection period.

[0010] Based on the collected environmental data, the environmental characteristic data within the detection period are calculated. The environmental characteristic data includes: the deviation rate of the current illuminance value of the photosensitive sensor in the detection period from the historical average value of the same period, the photovoltaic panel temperature, and the standard deviation of wind speed in the current detection period.

[0011] The environmental characteristic data within the detection period are preprocessed to remove non-stationary and outlier data from the original data.

[0012] First-order basic prediction calculation and second-order precision prediction calculation are performed on the preprocessed environmental feature data to obtain the second-order prediction results. Then, dynamic error correction is performed on the second-order prediction results to obtain the final prediction results, which are used as the prediction trend of output power fluctuation.

[0013] Furthermore, the preprocessing includes:

[0014] The prediction period is divided into multiple prediction sub-periods. The detection period is then divided equally according to the number of prediction sub-periods to obtain the detection sub-periods and the environmental feature data within each detection sub-period.

[0015] Check whether there are any anomalous data in the environmental feature data within each detection sub-cycle. If there are anomalous data, perform anomaly reconstruction, calculate the weighted average of the environmental feature data within the three adjacent detection sub-cycles of the anomalous data, and replace the anomalous data with the weighted average.

[0016] Then, a whitening process is performed. The whitening formula is as follows:

[0017] x'(k) = x(k) / max(x)

[0018] Where x(k) is the original data of each environmental feature data in the detection sub-cycle, max(x) is the maximum value of the environmental feature data in all detection sub-cycles, and x'(k) is the whitened data of each environmental feature data in the detection sub-cycle.

[0019] Furthermore, the first-order basic prediction calculation includes:

[0020] A smoothing trend curve for the detection period is generated based on the whitened data of environmental characteristic data within all detection sub-cycles, and then a prediction curve for the prediction period is generated based on the smoothing trend curve of the detection period.

[0021] The prediction curve of the prediction period is split based on the prediction sub-period to obtain the sub-period prediction curve of each prediction sub-period. The mean of the sub-period prediction curve is calculated to obtain the output power prediction value of each prediction sub-period.

[0022] After the first correction to the output power prediction value, the first-order basic prediction calculation result is obtained, as shown in the following formula:

[0023] P1=P0*(1+K env *H)

[0024] Where P0 is the predicted output power value for each prediction sub-cycle, P1 is the first-order predicted output power value after the first correction, and K... env H is the environmental impact factor, and H is the correction value. env When K is greater than 0, H is 0.01; when K is greater than 0, H is 0.01 env When H is less than 0, H is -0.01;

[0025] When the new energy power generation equipment is a photovoltaic power generation equipment, K env The calculation formula is as follows:

[0026] K env = *m+Kt*n

[0027] in, The deviation rate between the current detection sub-cycle illumination value and the historical average for the same period. Kt is the temperature influence coefficient, m is the deviation rate correction value, and n is the temperature influence correction value;

[0028] Kt=Tp / P

[0029] Where Tp is the photovoltaic panel temperature in the current detection sub-cycle, and P is the preset standard operating temperature of 25℃;

[0030] When the new energy power generation equipment is wind power generation equipment, K env The calculation formula is as follows:

[0031] K env =σV*r

[0032] Where σV is the standard deviation of wind speed in the current detection sub-cycle, and r is the standard deviation correction value.

[0033] Furthermore, the second-order precision prediction calculation includes:

[0034] The output difference between the predicted first-order output power value and the average actual output power value for each prediction sub-cycle is calculated using the following formula:

[0035] ε=│P actual -P1│

[0036] Where ε is the output difference, P actual This represents the average actual output power for the same period in history.

[0037] Based on the output difference, the first-order output power prediction value is corrected a second time, as shown in the following formula:

[0038] P2=P1+ε*λ

[0039] Where P2 is the second-order output power prediction value, and λ is the difference weighting coefficient. When ε < 5%, λ = 0.3; when 5% ≤ ε < 10%, λ = 0.6; and when 10% ≤ ε, λ = 0.9.

[0040] Furthermore, the dynamic error correction includes:

[0041] The formula for calculating the output difference between the predicted second-order output power and the average actual output power for the same historical period is as follows:

[0042] δ=|P2-P actual | / P actual ×100%

[0043] If the output difference δ > 10%, dynamic error correction is performed by calling the environmental impact factors corresponding to the standard deviations of the three most recent similar illumination values ​​and photovoltaic panel temperature or wind speed. The dynamic error correction formula is as follows:

[0044] P final =P2*(1+K env *H)

[0045] Among them, P final The result is the prediction after dynamic error correction.

[0046] The dynamic error correction prediction result is used as the output power prediction value for each prediction sub-cycle, and the output power prediction value for the entire prediction cycle is used as the output power fluctuation prediction trend. At the same time, the prediction confidence is calculated. If δ > risk threshold, a prediction risk warning is sent.

[0047] Furthermore, the output charging call instruction includes:

[0048] The energy transmission allocation weight for each energy storage unit is calculated using the following formula:

[0049] W final =(W SOC×0.4+W SOH ×0.2+W temp ×0.2+W grid ×0.2)

[0050] W SOC =1 - (SOC / 100)

[0051] W SOH =SOH / 100

[0052] W temp =1 - |T - 30| / 30

[0053] W grid =E / E'

[0054] Among them, W final Assign weights to the energy transmission of each energy storage unit, W SOC W is the charge state factor. SOH W is a health status factor. temp Here, T is the cell temperature, and W is the temperature factor. grid E is the grid demand factor, E is the energy output of the current energy storage unit to the grid, and E' is the average output of the current energy storage unit in the same historical period.

[0055] Based on the predicted output power value for each prediction sub-cycle, and combined with the current energy output from renewable energy power generation equipment to the grid, the energy stored by each renewable energy power generation equipment to the energy storage unit in each prediction sub-cycle is calculated using the following formula:

[0056] E available = (P final -P load )*

[0057] Among them, E available For energy storage, P load This refers to the energy output to the grid by current new energy power generation equipment. This is the time unit conversion factor, which is the value after converting the time of the predicted sub-cycle, and the unit is hours (h).

[0058] Based on the stored energy transmitted by the new energy power generation equipment and the energy transmission allocation weight, the energy value delivered by each new energy power generation equipment to the energy storage unit is calculated, as shown in the following formula:

[0059] E i_k =E available_k *W final_i

[0060] Among them, E i_k E represents the energy delivered to the i-th energy storage unit during the k-th prediction sub-cycle. available_kW represents the stored energy transmitted in the k-th predicted sub-cycle. final_i Assign weights to the energy transmission of the i-th energy storage unit;

[0061] According to the calculated E i_k The new energy power generation equipment receives the charging call instruction for each energy storage unit within the prediction period, and controls the power of energy delivered to the designated energy storage unit within the prediction period according to the charging call instruction.

[0062] Furthermore, after each prediction period ends, E is recalculated and updated. available_k and W final_i .

[0063] Preferably, within the regional coordination node, based on the SOC data collected by the multi-dimensional state monitoring module, when the SOC difference of energy storage units within the communication control area of ​​the regional coordination node exceeds a threshold, the high SOC energy storage unit is controlled to transfer energy to the low SOC energy storage unit through a bidirectional DC / DC converter, and the current balance is dynamically adjusted according to the SOC difference.

[0064] A distributed new energy storage device includes a service processor and a distributed memory. The service processor is connected to the memory, and the distributed memory stores a service self-management program configured to store machine-readable instructions. The service processor executes the service self-management program, and the instructions, when executed by the processor, realize the distributed new energy storage system as described above.

[0065] Compared to existing technologies, the beneficial effects of this application are as follows:

[0066] (1) This application collects multi-dimensional environmental data through a new energy power generation sensing module, optimizes the data quality through preprocessing, and then combines a multi-level prediction mechanism of first-order basic prediction, second-order precision prediction and dynamic error correction to improve the accuracy of output power fluctuation prediction trend.

[0067] (2) This application collects SOC, SOH and cell temperature in real time through a multi-dimensional state monitoring module. Based on the output power fluctuation prediction trend and the multi-dimensional state of the energy storage unit, the energy transmission allocation weight is determined by multi-factor weight calculation. Combined with the predicted energy storage demand of the sub-cycle, the energy transmission from the new energy power generation equipment to each energy storage unit is controlled to be accurately transmitted, thereby improving the utilization rate of energy storage resources. This enables the energy storage units to work in a balanced manner, avoids only a small number of energy storage units bearing the peak load for a long time, optimizes the energy storage performance and extends the service life. Attached Figure Description

[0068] In the attached diagram:

[0069] Figure 1 This is a schematic diagram of the system structure according to an embodiment of this application;

[0070] Figure 2 This is a schematic diagram of the device structure according to an embodiment of this application. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of this application, but not all embodiments. Generally, the components of the embodiments of this application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Example 1

[0072] like Figure 1 As shown, a distributed new energy storage system includes:

[0073] The new energy power generation sensing module is installed on new energy power generation equipment to sense new energy power generation and obtain the output power fluctuation prediction trend within the prediction period.

[0074] New energy power generation sensing includes:

[0075] Photosensitive sensors and temperature sensors are installed on photovoltaic power generation equipment, and wind speed sensors are installed on wind power generation equipment. The sensors are used to collect environmental data of the new energy power generation equipment during the detection period.

[0076] Based on the collected environmental data, the environmental characteristic data within the detection period are calculated. The environmental characteristic data includes: the deviation rate of the current illuminance value of the photosensitive sensor in the detection period from the historical average value of the same period, the photovoltaic panel temperature, and the standard deviation of wind speed in the current detection period.

[0077] The environmental characteristic data within the detection period are preprocessed to remove non-stationary and outlier data from the original data.

[0078] Preprocessing includes:

[0079] The prediction period is divided into multiple prediction sub-periods. The detection period is then divided equally according to the number of prediction sub-periods to obtain the detection sub-periods and the environmental feature data within each detection sub-period.

[0080] Check whether there are any anomalous data in the environmental feature data within each detection sub-cycle. If there are anomalous data, perform anomaly reconstruction, calculate the weighted average of the environmental feature data within the three adjacent detection sub-cycles of the anomalous data, and replace the anomalous data with the weighted average.

[0081] Then, a whitening process is performed. The whitening formula is as follows:

[0082] x'(k) = x(k) / max(x)

[0083] Where x(k) is the original data of each environmental feature data in the detection sub-cycle, max(x) is the maximum value of the environmental feature data in all detection sub-cycles, and x'(k) is the whitened data of each environmental feature data in the detection sub-cycle.

[0084] First-order basic prediction calculation and second-order precision prediction calculation are performed on the preprocessed environmental feature data to obtain the second-order prediction results. Then, dynamic error correction is performed on the second-order prediction results to obtain the final prediction results, which are used as the prediction trend of output power fluctuation.

[0085] First-order basic prediction calculations include:

[0086] A smoothing trend curve for the detection period is generated based on the whitened data of environmental characteristic data within all detection sub-cycles, and then a prediction curve for the prediction period is generated based on the smoothing trend curve of the detection period.

[0087] The prediction curve of the prediction period is split based on the prediction sub-period to obtain the sub-period prediction curve of each prediction sub-period. The mean of the sub-period prediction curve is calculated to obtain the output power prediction value of each prediction sub-period.

[0088] After the first correction to the output power prediction value, the first-order basic prediction calculation result is obtained, as shown in the following formula:

[0089] P1=P0*(1+K env *H)

[0090] Where P0 is the predicted output power value for each prediction sub-cycle, P1 is the first-order predicted output power value after the first correction, and K... env H is the environmental impact factor, and H is the correction value. env When K is greater than 0, H is 0.01; when K is greater than 0, H is 0.01 env When H is less than 0, H is -0.01;

[0091] When the new energy power generation equipment is a photovoltaic power generation equipment, K env The calculation formula is as follows:

[0092] K env = *m+Kt*n

[0093] in, The deviation rate between the current detection sub-cycle illumination value and the historical average for the same period. Kt is the temperature influence coefficient, m is the deviation rate correction value, and n is the temperature influence correction value;

[0094] Kt=Tp / P

[0095] Where Tp is the photovoltaic panel temperature in the current detection sub-cycle, and P is the preset standard operating temperature of 25℃;

[0096] When the new energy power generation equipment is wind power generation equipment, K env The calculation formula is as follows:

[0097] K env =σV*r

[0098] Where σV is the standard deviation of wind speed in the current detection sub-cycle, and r is the standard deviation correction value.

[0099] Second-order accuracy prediction calculations include:

[0100] The output difference between the predicted first-order output power value and the average actual output power value for each prediction sub-cycle is calculated using the following formula:

[0101] ε=│P actual -P1│

[0102] Where ε is the output difference, P actual This represents the average actual output power for the same period in history.

[0103] Based on the output difference, the first-order output power prediction value is corrected a second time, as shown in the following formula:

[0104] P2=P1+ε*λ

[0105] Where P2 is the second-order output power prediction value, and λ is the difference weighting coefficient. When ε < 5%, λ = 0.3; when 5% ≤ ε < 10%, λ = 0.6; and when 10% ≤ ε, λ = 0.9.

[0106] Dynamic error correction includes:

[0107] The formula for calculating the output difference between the predicted second-order output power and the average actual output power for the same historical period is as follows:

[0108] δ=|P2-P actual | / P actual ×100%

[0109] If the output difference δ > 10%, dynamic error correction is performed by calling the environmental impact factors corresponding to the standard deviations of the three most recent similar illumination values ​​and photovoltaic panel temperature or wind speed. The dynamic error correction formula is as follows:

[0110] P final =P2*(1+K env *H)

[0111] Among them, P final The result is the prediction after dynamic error correction.

[0112] The dynamic error correction prediction result is used as the output power prediction value for each prediction sub-cycle, and the output power prediction value for the entire prediction cycle is used as the output power fluctuation prediction trend. At the same time, the prediction confidence is calculated. If δ > risk threshold, a prediction risk warning is sent.

[0113] A multi-dimensional status monitoring module is installed on the energy storage unit to collect the SOC, SOH and cell temperature of the energy storage unit in real time;

[0114] The multi-dimensional status monitoring module is equipped with a unit protection module, which is used to trigger local disconnection when the measured cell temperature difference is greater than the threshold or the SOC deviates from the threshold range, so as to avoid the spread of the fault.

[0115] The energy storage unit contains an energy storage battery module, a buffer capacitor module, and a vanadium redox flow battery module;

[0116] Energy storage battery modules are responsible for basic energy storage;

[0117] The buffer capacitor module is used to absorb or release short-term impact loads when the power generation of new energy power generation equipment fluctuates.

[0118] Vanadium redox flow battery modules are used for day and night energy storage, extending the lifespan of energy storage battery modules.

[0119] The buffer capacitor module handles short-term peak loads, the energy storage battery module handles medium-term loads, and the vanadium redox flow battery module handles long-term energy storage.

[0120] The regional coordination node establishes communication connections with multiple new energy power generation sensing modules and multi-dimensional status monitoring modules within the communication distance. It is used to receive data collected by the new energy power generation sensing modules and multi-dimensional status monitoring modules at preset frequencies, and output charging call commands based on the output power fluctuation prediction trend and SOC, SOH and cell temperature to control the power of new energy power generation equipment to transmit energy to different energy storage units.

[0121] SOC stands for State of Charge, and SOH stands for State of Health.

[0122] Output charging command instructions include:

[0123] The energy transmission allocation weight for each energy storage unit is calculated using the following formula:

[0124] W final =(W SOC ×0.4+W SOH ×0.2+W temp ×0.2+W grid ×0.2)

[0125] W SOC=1 - (SOC / 100)

[0126] W SOH =SOH / 100

[0127] W temp =1 - |T - 30| / 30

[0128] W grid =E / E'

[0129] Among them, W final Assign weights to the energy transmission of each energy storage unit, W SOC W is the charge state factor. SOH W is a health status factor. temp Here, T is the cell temperature, and W is the temperature factor. grid E is the grid demand factor, E is the energy output of the current energy storage unit to the grid, and E' is the average output of the current energy storage unit in the same historical period.

[0130] Based on the predicted output power value for each prediction sub-cycle, and combined with the current energy output from renewable energy power generation equipment to the grid, the energy stored by each renewable energy power generation equipment to the energy storage unit in each prediction sub-cycle is calculated using the following formula:

[0131] E available = (P final -P load )*

[0132] Among them, E available For energy storage, P load This refers to the energy output to the grid by current new energy power generation equipment. This is a time unit conversion factor, which is the value after converting the time of the prediction sub-period, in hours (h). When the prediction sub-period is 3 minutes... = ( h;

[0133] Based on the stored energy transmitted by the new energy power generation equipment and the energy transmission allocation weight, the energy value delivered by each new energy power generation equipment to the energy storage unit is calculated, as shown in the following formula:

[0134] E i_k =E available_k *W final_i

[0135] Among them, E i_k E represents the energy delivered to the i-th energy storage unit during the k-th prediction sub-cycle. available_k W represents the stored energy transmitted in the k-th predicted sub-cycle. final_i Assign weights to the energy transmission of the i-th energy storage unit;

[0136] According to the calculated E i_k The new energy power generation equipment receives the charging call instruction for each energy storage unit within the prediction period, and controls the power of energy delivered to the designated energy storage unit within the prediction period according to the charging call instruction.

[0137] At the end of each forecast period, E is recalculated and updated. available_k and W final_i .

[0138] The power transmitted by regional coordination node control of new energy power generation equipment to different energy storage units includes:

[0139] State of charge (SOC) takes precedence over state of oxygen (SOH), and state of oxygen (SOH) takes precedence over cell temperature.

[0140] Low-SOC energy storage units have higher priority than high-SOC energy storage units;

[0141] High SOH energy storage units have higher priority than low SOH energy storage units;

[0142] Low cell temperature energy storage units have a higher priority than high cell temperature energy storage units.

[0143] Within the regional coordination node, based on the SOC data collected by the multi-dimensional status monitoring module, when the SOC difference between energy storage units within the communication control area of ​​the regional coordination node exceeds a threshold, the high SOC energy storage unit is controlled to transfer energy to the low SOC energy storage unit through a bidirectional DC / DC converter, and the current balance is dynamically adjusted according to the SOC difference. Example 2

[0144] like Figure 2 As shown, a distributed new energy storage device includes a service processor and a distributed memory. The service processor is connected to the memory, and the distributed memory stores a service self-management program configured to store machine-readable instructions. The service processor executes the service self-management program, and the instructions, when executed by the processor, implement the distributed new energy storage system as described in Embodiment 1.

[0145] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications, improvements, and substitutions without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. A distributed new energy storage system, characterized in that, include: A new energy power generation sensing module, installed on new energy power generation equipment, is used to sense new energy power generation and obtain the predicted trend of output power fluctuations within a prediction period. New energy power generation sensing includes: Photosensitive sensors and temperature sensors are installed on photovoltaic power generation equipment, and wind speed sensors are installed on wind power generation equipment. The sensors are used to collect environmental data of the new energy power generation equipment during the detection period. Based on the collected environmental data, the environmental characteristic data within the detection period are calculated. The environmental characteristic data includes: the deviation rate of the current illuminance value of the photosensitive sensor in the detection period from the historical average value of the same period, the photovoltaic panel temperature, and the standard deviation of wind speed in the current detection period. The environmental characteristic data within the detection period are preprocessed to remove non-stationary and outlier data from the original data. First-order basic prediction calculation and second-order precision prediction calculation are performed on the preprocessed environmental feature data to obtain the second-order prediction result. Then, dynamic error correction is performed on the second-order prediction result to obtain the final prediction result, which is used as the output power fluctuation prediction trend. A multi-dimensional status monitoring module is installed on the energy storage unit to collect the SOC, SOH and cell temperature of the energy storage unit in real time; The regional coordination node receives data collected by the new energy power generation sensing module and the multi-dimensional status monitoring module. Based on the output power fluctuation prediction trend and SOC, SOH, and cell temperature, it outputs charging call commands to control the power of the new energy power generation equipment to transfer energy to different energy storage units. The charging call commands include: The energy transmission allocation weight for each energy storage unit is calculated using the following formula: IN final =(W SOC ×0.4+W SOH ×0.2+W temp ×0.2+W grid ×0.2) W SOC =1-(SOC / 100) W SOH =SOH / 100 W temp =1-|T-30| / 30 W grid =Y / Y' Among them, W final Assign weights to the energy transmission of each energy storage unit, W SOC W is the charge state factor. SOH W is a health status factor. temp Here, T is the cell temperature, and W is the temperature factor. grid E is the grid demand factor, E is the energy output of the current energy storage unit to the grid, and E' is the average output of the current energy storage unit in the same historical period. Based on the predicted output power value for each prediction sub-cycle, and combined with the current energy output from renewable energy power generation equipment to the grid, the energy stored by each renewable energy power generation equipment to the energy storage unit in each prediction sub-cycle is calculated using the following formula: E available =(P final -P load )*Δt Among them, E available For energy storage, P load The energy output from the current new energy power generation equipment to the grid is represented by Δt, which is the time unit conversion factor, representing the value after converting the predicted sub-cycle time, with the unit being hours (h). Based on the stored energy transmitted by the new energy power generation equipment and the energy transmission allocation weight, the energy value delivered by each new energy power generation equipment to the energy storage unit is calculated as follows: HAVE BEEN i_k =E available_k *W final_i Among them, E i_k E represents the energy delivered to the i-th energy storage unit during the k-th prediction sub-cycle. available_k W represents the stored energy transmitted in the k-th predicted sub-cycle. final_i Assign weights to the energy transmission of the i-th energy storage unit; According to the calculated E i_k The new energy power generation equipment receives the charging call instruction for each energy storage unit within the prediction period, and controls the power of energy delivered to the designated energy storage unit within the prediction period according to the charging call instruction.

2. The distributed new energy storage system according to claim 1, characterized in that, The preprocessing includes: The prediction period is divided into multiple prediction sub-periods. The detection period is then divided equally according to the number of prediction sub-periods to obtain the detection sub-periods and the environmental feature data within each detection sub-period. Check if there are any anomalous data in the environmental feature data within each detection sub-cycle. If there are anomalous data, perform anomaly reconstruction, calculate the weighted average of the environmental feature data within the three adjacent detection sub-cycles of the anomalous data, and replace the anomalous data with the weighted average. Then, a whitening process is performed. The whitening formula is as follows: x'(k) = x(k) / max(x) Where x(k) is the original data of each environmental feature data in the detection sub-cycle, max(x) is the maximum value of the environmental feature data in all detection sub-cycles, and x'(k) is the whitened data of each environmental feature data in the detection sub-cycle.

3. A distributed new energy storage system according to claim 1, characterized in that, The first-order basic prediction calculation includes: A smoothing trend curve for the detection period is generated based on the whitened data of environmental characteristic data within all detection sub-cycles, and then a prediction curve for the prediction period is generated based on the smoothing trend curve of the detection period. The prediction curve of the prediction period is split based on the prediction sub-period to obtain the sub-period prediction curve of each prediction sub-period. The mean of the sub-period prediction curve is calculated to obtain the output power prediction value of each prediction sub-period. After the first correction to the output power prediction value, the first-order basic prediction calculation result is obtained, as shown in the following formula: P1=P0*(1+K env *H) Where P0 is the predicted output power value for each prediction sub-cycle, P1 is the first-order predicted output power value after the first correction, and K... env H is the environmental impact factor, and H is the correction value. env When K is greater than 0, H is 0.01; when K is greater than 0, H is 0.01 env When H is less than 0, H is -0.01; When the new energy power generation equipment is a photovoltaic power generation equipment, K env The calculation formula is as follows: K env =ΔG*m+Kt*n Wherein, ΔG is the deviation rate between the current detection sub-cycle illumination value and the historical average value for the same period, Kt is the temperature influence coefficient, m is the deviation rate correction value, and n is the temperature influence correction value; Kt = Tp / P Where Tp is the photovoltaic panel temperature in the current detection sub-cycle, and P is the preset standard operating temperature of 25℃; When the new energy power generation equipment is wind power generation equipment, K env The calculation formula is as follows: K env =σV*r Where σV is the standard deviation of wind speed in the current detection sub-cycle, and r is the standard deviation correction value.

4. A distributed new energy storage system according to claim 3, characterized in that, The second-order accuracy prediction calculation includes: The output difference between the predicted first-order output power value and the average actual output power value for each prediction sub-cycle is calculated using the following formula: ε=│P actual -P1│ Where ε is the output difference, P actual This represents the average actual output power for the same period in history. Based on the output difference, the first-order output power prediction value is corrected a second time, as shown in the following formula: P2=P1+ε*λ Where P2 is the second-order output power prediction value, and λ is the difference weighting coefficient. When ε < 5%, λ = 0.3; when 5% ≤ ε < 10%, λ = 0.6; and when 10% ≤ ε, λ = 0.

9.

5. A distributed new energy storage system according to claim 4, characterized in that, The dynamic error correction includes: The formula for calculating the output difference between the predicted second-order output power and the average actual output power for the same historical period is as follows: δ=|P2-P actual | / P actual ×100% If the output difference δ > 10%, dynamic error correction is performed by calling the environmental impact factors corresponding to the standard deviations of the three most recent similar illumination values ​​and photovoltaic panel temperature or wind speed. The dynamic error correction formula is as follows: P final =P2*(1+K env *H) Among them, P final The result is the prediction after dynamic error correction. The dynamic error correction prediction result is used as the output power prediction value for each prediction sub-cycle, and the output power prediction value for the entire prediction cycle is used as the output power fluctuation prediction trend. At the same time, the prediction confidence is calculated. If δ > risk threshold, a prediction risk warning is sent.

6. A distributed new energy storage system according to claim 1, characterized in that, At the end of each prediction period, the updated E is recalculated. available_k and W final_i .

7. A distributed new energy storage system according to claim 1, characterized in that, Within the regional coordination node, based on the SOC data collected by the multi-dimensional state monitoring module, when the SOC difference between energy storage units within the communication control area of ​​the regional coordination node exceeds a threshold, the node controls the high-SOC energy storage unit to transfer energy to the low-SOC energy storage unit through a bidirectional DC / DC converter, and dynamically adjusts the current balance according to the SOC difference.

8. A distributed new energy storage device, characterized in that, The device includes a service processor and a distributed memory. The service processor is connected to the memory. The distributed memory stores a service self-management program configured to store machine-readable instructions. The service processor executes the service self-management program. When the instructions are executed by the processor, they implement the distributed new energy storage system as described in claim 1.

Images