Energy storage multi-module collaborative energy conversion and efficiency enhancement system and method

Abstract

The application relates to the technical field of multi-source flexible grid connection, and discloses a storage multi-module cooperative energy conversion efficiency increasing system and method, the method comprising the following steps: high-frequency sampling of load change data at a load end, and collecting operation state data of each energy conversion unit; time scale analysis of energy change characteristics at the load end, and calculation of dynamic characteristic strength representing rapid energy change; judging whether there is significant energy fluctuation at the load end; if yes, constructing a dynamic response strategy for the load end, comprising the following steps: calculation of dynamic response adaptation indexes of each energy conversion unit; screening of response adjustment units, and calculation of adjustment task weights of each response adjustment unit; based on the corresponding adjustment task weights, controlling each response adjustment unit to perform dynamic response adjustment on the load end. The application can improve the overall conversion efficiency and operation stability of a solid-state transformer in the scene of flexible grid connection and multi-load direct connection.

Description

Technical Field

[0001] This application relates to the technical field of multi-source flexible grid connection, specifically to a multi-module collaborative energy conversion efficiency enhancement system and method for energy storage. Background Technology

[0002] In flexible grid-connected solid-state transformers designed for direct access to multiple DC sources such as photovoltaics, wind power, and energy storage, the system typically consists of multiple DC-DC, DC-IC, and ICDC energy conversion modules working collaboratively to achieve direct-connection-free and efficient grid connection. However, in actual operation, the load side, especially the ICDC-connected load or grid interface, exhibits significant energy fluctuation characteristics, and different energy conversion units show significant differences in energy regulation margin and rapid power response. Existing technologies often employ uniform current sharing or power allocation strategies, failing to differentiate the dynamic response capabilities of different energy conversion units under different energy fluctuation conditions and time scales. This leads to frequent inefficient high-frequency modulation by the energy conversion units, generating implicit losses during collaboration; thus, it restricts conversion efficiency and collaborative stability, making it difficult to support the high-conversion-rate flexible grid-connected operation requirements.

[0003] For example, patent application CN115603356A discloses a new energy DC collection and transmission system and its control method based on a DC transformer. The system includes a renewable energy collection DC grid, a DC transformer, a long-distance DC transmission system, and multiple renewable energy grid-connected devices. The renewable energy grid-connected devices receive externally input renewable energy signals, perform signal transformation, and then input the signals to the renewable energy collection DC grid for signal collection. The collected signals are then voltage-transformed by the DC transformer and output through the long-distance DC transmission system. This solution facilitates the access of new energy power plants of different capacities and has good compatibility and scalability. However, it still suffers from the problem mentioned in the background of this application: hidden losses occur during multi-source coordination, which restricts conversion efficiency and coordination stability.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the overall background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to overcome the shortcomings of the prior art and provide a multi-module collaborative energy conversion efficiency enhancement system and method for energy storage, thereby improving the overall conversion efficiency and operational stability of solid-state transformers in flexible grid connection and multi-load direct connection scenarios.

[0006] To solve the above-mentioned technical problems, this application provides the following technical solution:

[0007] On the one hand, this application provides a method for enhancing energy conversion efficiency through multi-module collaborative energy storage, comprising the following steps:

[0008] The load change data at the load end is sampled at high frequency, and the operating status data of each energy conversion unit is collected;

[0009] Based on the load change data, the energy change characteristics at the load end are analyzed on a time scale, and the intensity of dynamic characteristics characterizing rapid energy changes is calculated.

[0010] Based on the intensity of the dynamic characteristics, determine whether there are significant energy fluctuations at the load end; if so, construct a dynamic response strategy for the load end, including:

[0011] The dynamic response adaptation index of each energy conversion unit is calculated based on the aforementioned operating status data;

[0012] Response adjustment units are selected based on the dynamic response adaptation index, and the adjustment task weight of each response adjustment unit is calculated.

[0013] Based on the corresponding adjustment task weights, each response adjustment unit is controlled to dynamically adjust the load side.

[0014] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the load change data includes load voltage and load current; based on the load change data, a time-scale analysis of the energy change characteristics at the load end is performed, specifically including:

[0015] The instantaneous power at the load terminal at each moment is calculated based on the load voltage and load current, and the instantaneous power sequence at the load terminal is constructed.

[0016] Set up a multi-scale analysis window; perform time-scale analysis based on the instantaneous power sequence and the multi-scale analysis window, and construct an energy change sequence;

[0017] The intensity of dynamic features characterizing rapid energy changes is calculated based on the energy change sequence;

[0018] The multi-scale analysis window includes a first-scale window; the energy change sequence includes a first energy sequence; the method for constructing the first energy sequence is as follows: based on the instantaneous power sequence and the first-scale window, the first-scale power at each moment is calculated, and the first energy sequence is constructed; the first energy sequence is a time series of the first-scale power; the method for calculating the first-scale power at any moment is as follows:

[0019] Mark any time point as the first target time point; extract the instantaneous power sequence through a first scale window, with the endpoint of the first scale window being the first target time point; calculate the mean of the instantaneous power at each time point within the first scale window as the first smoothed power; calculate the difference between the instantaneous power at the first target time point and the first smoothed power as the first scale power at the first target time point.

[0020] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, wherein: the multi-scale analysis window further includes a second-scale window, and the length of the first-scale window is less than the length of the second-scale window; the energy change sequence further includes a second energy sequence;

[0021] The method for constructing the second energy sequence is as follows: Based on the instantaneous power sequence and the second scale window, the second scale power at each moment is calculated, and the second energy sequence is constructed; the second energy sequence is a time series of the second scale power; the method for calculating the second scale power at any moment is as follows:

[0022] Mark any time point as the second target time point; extract the instantaneous power sequence through a second scale window, with the endpoint of the second scale window being the second target time point; calculate the mean of the instantaneous power at each time point within the second scale window as the second smoothed power; calculate the difference between the first smoothed power and the second smoothed power at the second target time point as the second scale power at the second target time point.

[0023] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the dynamic characteristic intensity includes a first characteristic intensity and a second characteristic intensity; the method for calculating the first characteristic intensity is as follows: setting a load detection window, the load detection window containing the most recent M moments; M is a positive integer;

[0024] Based on the load detection window, the instantaneous power of the most recent M moments is extracted from the instantaneous power sequence; the absolute value of the instantaneous power of the M moments within the load detection window is taken and the mean is calculated as a reference characteristic index;

[0025] Based on the load detection window, the first-scale power of the most recent M moments is extracted from the first energy sequence; the absolute value of the first-scale power of the M moments within the load detection window is taken and the mean is calculated as the first feature index.

[0026] The ratio of the first feature index to the reference feature index is calculated and used as the first feature intensity.

[0027] The method for calculating the intensity of the second feature is as follows:

[0028] Based on the load detection window, the second-scale power of the most recent M moments is extracted from the second energy sequence; the absolute value of the second-scale power of the M moments within the load detection window is taken and the mean is calculated as the second feature index; the ratio of the second feature index to the reference feature index is calculated as the second feature intensity.

[0029] The following methods can be used to determine whether there are significant energy fluctuations at the load end:

[0030] Set a first intensity threshold and a second intensity threshold; if at least one of the following conditions is met, namely, the first characteristic intensity is greater than the first intensity threshold and the second characteristic intensity is greater than the second intensity threshold, then there is a significant energy fluctuation at the load end.

[0031] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the operating status data of any energy conversion unit includes input power and output power; the method for calculating the dynamic response adaptation index of any energy conversion unit is as follows:

[0032] A power supply detection window is set; the power supply detection window contains the most recent N times; N is a positive integer;

[0033] Extract the input power of the energy conversion unit at N time points within the power supply detection window, and calculate the standard deviation as the input index of the energy conversion unit; assign a value to the first fit of the energy conversion unit based on the input index;

[0034] Extract the output power of the energy conversion unit at N times within the power supply detection window, and calculate the average value as the output index of the energy conversion unit;

[0035] Obtain the rated output power of the energy conversion unit; assign a value based on the rated output power and the second fit of the output index of the energy conversion unit;

[0036] The product of the first fit and the second fit is calculated and used as the dynamic response fit index of the energy conversion unit.

[0037] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the method for selecting the response adjustment unit is as follows:

[0038] The energy conversion units are sorted in descending order of dynamic response adaptation index; the energy conversion units ranked first are selected as response adjustment units according to a preset selection ratio.

[0039] The method for calculating the adjustment task weight of any response adjustment unit is as follows: assign a value to the adjustment task weight of each response adjustment unit based on the dynamic response adaptation index; the adjustment task weight of any response adjustment unit is positively correlated with the corresponding dynamic response adaptation index, and the sum of the adjustment task weights of all response adjustment units is 1.

[0040] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the dynamic response adjustment of the load end specifically includes:

[0041] Based on the corresponding dynamic response adaptation index and rated output power, the response adjustment period and power adjustment threshold are set for each response adjustment unit.

[0042] An energy deviation threshold is set for each response adjustment unit. At the beginning of each response adjustment cycle, the energy deviation of the response adjustment unit is calculated for any response adjustment unit. If the absolute value of the energy deviation is greater than the corresponding energy deviation threshold, the energy deviation is suppressed by adjusting the output power of the response adjustment unit, and the adjustment amount of the output power is less than or equal to the corresponding power adjustment threshold.

[0043] As a preferred embodiment of the energy storage multi-module collaborative energy conversion efficiency enhancement method described in this application, the method for setting the response adjustment period for any response adjustment unit is as follows: setting a benchmark adjustment period; assigning values ​​to the length of the response adjustment period of each response adjustment unit based on the benchmark adjustment period and the dynamic response adaptation index, wherein the length of the response adjustment period of any response adjustment unit is negatively correlated with the corresponding dynamic response adaptation index;

[0044] The method for setting the power regulation threshold for any response regulation unit is as follows: assign a value to the power regulation threshold of each response regulation unit based on the rated output power and the dynamic response adaptation index, and the power regulation threshold of any response regulation unit is positively correlated with the corresponding rated output power and dynamic response adaptation index.

[0045] The method for calculating the energy deviation of any response regulation unit is as follows: the total energy deviation at the load end is calculated based on the load voltage at the start of the response regulation cycle and at the start of the previous response regulation cycle; the energy deviation of the response regulation unit is the product of the total energy deviation and the corresponding regulation task weight.

[0046] Secondly, this application provides an energy storage multi-module collaborative energy conversion efficiency enhancement system for realizing the energy storage multi-module collaborative energy conversion efficiency enhancement method as described in the first aspect, including a flexible grid-connected solid-state transformer and a control subsystem;

[0047] The flexible grid-connected solid-state transformer includes an input module, an energy conversion module, and an output module; wherein, the input module includes multiple DC access units; any DC access unit is used to connect to the current output from an external DC source;

[0048] The energy conversion module includes multiple energy conversion units; each energy conversion unit is connected to a DC input unit and is used to convert the current input to the DC input unit into energy.

[0049] The output module is used to integrate the current output by the energy conversion unit and supply power to the load.

[0050] The control subsystem includes a data acquisition module, a data processing module, and a control module; wherein, the data acquisition module is used to acquire load change data at the load end and operating status data of each energy conversion unit;

[0051] The data processing module calculates the dynamic characteristic intensity of the energy at the load end based on the load change data, and calculates the dynamic response adaptation index of each energy conversion unit based on the operating status data.

[0052] The control module determines whether there is significant energy fluctuation at the load end based on the dynamic characteristic intensity; if so, the control module selects response adjustment units based on the dynamic response adaptation index and controls each response adjustment unit to perform dynamic response adjustment at the load end.

[0053] Compared with the prior art, the beneficial effects achieved by this application are as follows:

[0054] This application performs dynamic characteristic analysis of load changes at the time scale level, combines the operating status of each energy conversion unit, selectively activates energy conversion units suitable for dynamic response to participate in load regulation, and achieves multi-unit coordinated response by adjusting task weights. Thus, without relying on fixed module division of labor or complex control structures, it achieves efficient suppression of load energy fluctuations, improves the system's dynamic adaptability, operational stability, and overall energy conversion efficiency in multi-energy access scenarios, and reduces system losses caused by unnecessary high-frequency modulation during multi-source coordination, thereby improving the overall conversion efficiency and operational stability of solid-state transformers in flexible grid connection and multi-load direct connection scenarios. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0056] Figure 1 A flowchart of the energy storage multi-module collaborative energy conversion efficiency enhancement method provided in this application;

[0057] Figure 2 This is a schematic diagram of the structure of the multi-module collaborative energy conversion and efficiency enhancement system for energy storage provided in this application. Detailed Implementation

[0058] The technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof. In the absence of conflict, the embodiments and technical features in the embodiments can be combined with each other.

[0059] Example 1

[0060] This embodiment introduces a method for enhancing energy conversion efficiency through multi-module collaborative energy storage, referring to... Figure 1 The method includes the following steps:

[0061] The load change data at the load end is sampled at high frequency, and the operating status data of each energy conversion unit is collected;

[0062] The load change data includes load voltage and load current; optionally, the load change data can be sampled at high frequency as follows: simultaneously collect the load voltage and load current at the load end with the same high frequency sampling period, such as 2ms; record the load voltage and load current at each sampling point and mark the corresponding timestamp.

[0063] In this embodiment, the operating status data of any energy conversion unit includes input power and output power. The input power is the power supplied to the energy conversion unit by an external DC source, and the output power is the power supplied by the energy conversion unit to the load.

[0064] Based on the load change data, the energy change characteristics at the load end are analyzed on a time scale, and the intensity of dynamic characteristics characterizing rapid energy changes is calculated.

[0065] The time-scale analysis of the energy change characteristics at the load end specifically includes:

[0066] The instantaneous power of the load terminal at each moment is calculated based on the load voltage and load current, and the instantaneous power sequence of the load terminal is constructed; the instantaneous power sequence is a time series of the instantaneous power of the load terminal; the instantaneous power of the load terminal at any moment is the product of the load voltage and load current at the corresponding moment.

[0067] Set up a multi-scale analysis window; perform time-scale analysis based on the instantaneous power sequence and the multi-scale analysis window, and construct an energy change sequence;

[0068] The multi-scale analysis window includes a first-scale window and a second-scale window, wherein the length of the first-scale window is shorter than the length of the second-scale window; the energy change sequence includes a first energy sequence and a second energy sequence.

[0069] The first scale window is used to extract high-frequency features of load-side energy changes, and has the smallest window size. The second scale window is used to extract mid-to-high-frequency features of load-side energy changes, and has a smaller window size, but larger than the first scale window. Those skilled in the art can set the lengths of the first and second scale windows based on actual needs; for example, the length of the first scale window can be 20ms, and the length of the second scale window can be 100ms.

[0070] The method for constructing the first energy sequence is as follows: Based on the instantaneous power sequence and the first scale window, the first scale power at each time moment is calculated, and the first energy sequence is constructed; the first energy sequence is a time series of the first scale power; the method for calculating the first scale power at any time moment is as follows:

[0071] Mark any time point as the first target time point; extract the instantaneous power sequence through a first scale window, with the endpoint of the first scale window being the first target time point; calculate the mean of the instantaneous power at each time point within the first scale window as the first smoothed power; calculate the difference between the instantaneous power at the first target time point and the first smoothed power as the first scale power at the first target time point.

[0072] The first-scale power represents the high-frequency jitter component in the load segment power at the corresponding moment. Taking a first-scale window of 20ms as an example, the first smoothed power is the average power level within the most recent first-scale window relative to the first target moment. It is the load power after smoothing and suppressing the high-frequency jitter component. It can reflect the power fluctuation at the 20ms scale, but cannot reflect the power fluctuation at a higher precision. When observing at the highest precision scale, i.e., the original high-frequency sampling scale, there are faster power fluctuations in the instantaneous power sequence. Subtracting the first smoothed power from the instantaneous power can eliminate the overall trend at the 20ms observation scale, obtaining the high-frequency jitter component at the observation scale of several milliseconds, i.e., the first-scale power.

[0073] The method for constructing the second energy sequence is as follows: Based on the instantaneous power sequence and the second scale window, the second scale power at each moment is calculated, and the second energy sequence is constructed; the second energy sequence is a time series of the second scale power; the method for calculating the second scale power at any moment is as follows:

[0074] Mark any time point as the second target time point; extract the instantaneous power sequence through a second scale window, with the endpoint of the second scale window being the second target time point; calculate the mean of the instantaneous power at each time point within the second scale window as the second smoothed power; calculate the difference between the first smoothed power and the second smoothed power at the second target time point as the second scale power at the second target time point.

[0075] The second-scale power represents the mid-to-high frequency jitter component in the load power at the corresponding time. Taking a second-scale window of 100ms as an example, the second smoothed power is the average power level within the most recent second-scale window relative to the second target time. It is the load power after smoothing and suppressing the high-frequency and mid-to-high frequency jitter components, and it can reflect the power fluctuation at the 100ms scale. Subtracting the second smoothed power from the first smoothed power eliminates the overall trend at the 100ms observation scale, yielding the mid-to-high frequency jitter component at the 20ms observation scale, i.e., the second-scale power.

[0076] The intensity of dynamic features characterizing rapid energy changes is calculated based on the energy change sequence.

[0077] The dynamic feature intensity includes a first feature intensity and a second feature intensity;

[0078] The method for calculating the intensity of the first feature is as follows:

[0079] Set a load detection window, which contains the most recent M time points; M is a positive integer.

[0080] Based on the load detection window, the instantaneous power of the most recent M moments is extracted from the instantaneous power sequence; the absolute value of the instantaneous power of the M moments within the load detection window is taken and the mean is calculated as a reference characteristic index;

[0081] Based on the load detection window, the first-scale power of the most recent M moments is extracted from the first energy sequence; the absolute value of the first-scale power of the M moments within the load detection window is taken and the mean is calculated as the first feature index.

[0082] The ratio of the first feature index to the reference feature index is calculated and used as the first feature intensity.

[0083] The method for calculating the intensity of the second feature is as follows:

[0084] Based on the load detection window, the second-scale power of the most recent M moments is extracted from the second energy sequence; the absolute value of the second-scale power of the M moments within the load detection window is taken and the mean is calculated as the second feature index.

[0085] The ratio of the second feature index to the reference feature index is calculated as the second feature strength.

[0086] Based on the intensity of the dynamic characteristics, determine whether there are significant energy fluctuations at the load end;

[0087] The following methods can be used to determine whether there are significant energy fluctuations at the load end:

[0088] Set a first intensity threshold and a second intensity threshold; if at least one of the following conditions is met, namely, the first characteristic intensity is greater than the first intensity threshold and the second characteristic intensity is greater than the second intensity threshold, then there is a significant energy fluctuation at the load end.

[0089] The first intensity characteristic reflects the intensity of transient, sharp high-frequency energy fluctuations at the load end, while the second intensity characteristic reflects the intensity of high-frequency energy fluctuations at the load end that last for a relatively longer period. Taking a 380V DC bus load in a data center as an example, when observed on a timescale of several seconds or longer, its instantaneous power can be maintained at 9.5-10.5kW for a long period, corresponding to its relatively stable business load. If observed on a timescale of tens of milliseconds, power fluctuations of hundreds to thousands of watts will appear, which is common in services such as server cluster task scheduling and fan PWM group response. For example, if there is a batch processing every 100ms, the second energy sequence will show quasi-periodic fluctuations. If observed on a timescale of several milliseconds, spikes or jitters of tens to hundreds of watts will appear, which is common in rapid adjustments such as CPU transient loads. For example, the instantaneous jump in the load of AI inference leads to high-frequency energy fluctuations, producing short spikes. This application measures the intensity and duration of these power fluctuations, spikes, and jitters by calculating the intensity of dynamic characteristics. If the intensity of dynamic characteristics is too large, the energy fluctuations at the load end cannot be ignored and intervention is required to maintain power supply stability. Those skilled in the art can set specific values ​​for the first and second intensity thresholds based on actual needs. For example, a first intensity threshold of 0.02 and a second intensity threshold of 0.05 would be more sensitive to the identification of the first characteristic intensity representing rapid changes in load-side energy, allowing for earlier identification of transient fluctuations; while the identification of the second characteristic intensity representing moderate to rapid changes in load-side energy would require more significant energy fluctuations to reduce unnecessary dynamic adjustment triggering.

[0090] If there are significant energy fluctuations at the load end, a dynamic response strategy is constructed for the load end; including:

[0091] The dynamic response adaptation index of each energy conversion unit is calculated based on the aforementioned operating status data;

[0092] The method for calculating the dynamic response adaptation index of any energy conversion unit is as follows:

[0093] A power supply detection window is set; the power supply detection window contains the most recent N times; N is a positive integer;

[0094] Extract the input power of the energy conversion unit at N time points within the power supply detection window, and calculate the standard deviation as the input index of the energy conversion unit; assign a value to the first fit of the energy conversion unit based on the input index;

[0095] Optionally, the first fit degree can be assigned a value as follows: An input reference index is set based on actual needs; the first fit degree is assigned as 1 minus the ratio of the input index to the input reference index, i.e., the larger the input index (the closer it is to the income reference index), the lower the first fit degree. A higher first fit degree indicates a more stable power supply on the input side, such as when the energy conversion unit is connected to a steady-state DC source like an energy storage cabinet, making it suitable for rapid response adjustments. A higher first fit degree also indicates greater power fluctuations on the input side, such as when the current input comes from wind power generation under turbulent conditions; in such cases, the energy conversion unit is not suitable for rapid response adjustment tasks, otherwise, it may easily introduce new energy fluctuations.

[0096] Extract the output power of the energy conversion unit at N times within the power supply detection window, and calculate the average value as the output index of the energy conversion unit;

[0097] Obtain the rated output power of the energy conversion unit; assign a value based on the rated output power and the second fit of the output index of the energy conversion unit;

[0098] Alternatively, the formula for calculating the second fitness is as follows:

[0099] ;

[0100] Where s represents the second fitness, and P represents the output metric. This represents the rated output power. Based on the above formula, the closer the output specification is to 0.5 times the rated output power, the greater the second fit. When the output specification is too small, the energy supply on the input side is insufficient, lacking the energy margin available for dynamic response adjustment. When the output specification is too large, the energy conversion unit is already operating at near full load, and its dynamic space for adjusting energy fluctuations is very small, making it difficult to effectively participate in correcting energy fluctuations at the load end.

[0101] The product of the first fit and the second fit is calculated and used as the dynamic response fit index of the energy conversion unit.

[0102] Response adjustment units are selected based on the dynamic response adaptation index, and the adjustment task weight of each response adjustment unit is calculated.

[0103] The method for filtering the response adjustment unit is as follows:

[0104] The energy conversion units are sorted in descending order of dynamic response adaptation index; the energy conversion units ranked higher are selected as response adjustment units according to a preset selection ratio.

[0105] Those skilled in the art can set the selection ratio based on actual needs, for example, 20%, that is, selecting the top 20% of energy conversion units with the highest dynamic response adaptation index as response adjustment units. The remaining energy conversion units that are not selected as response adjustment units mainly undertake the basic power supply part of the load end, and will not respond frequently to the instantaneous fluctuations of the load end, thereby ensuring the overall power supply efficiency and reliability.

[0106] The method for calculating the adjustment task weight of any response adjustment unit is as follows: the adjustment task weight of each response adjustment unit is assigned a value based on the dynamic response adaptation index, and the adjustment task weight of any response adjustment unit is positively correlated with the corresponding dynamic response adaptation index, and the sum of the adjustment task weights of all response adjustment units is 1.

[0107] Optionally, the sum of the dynamic response adaptation indices of all response regulation units is calculated as a normalization coefficient; the regulation task weight of any response regulation unit is the ratio of its dynamic response adaptation index to the normalization coefficient. When multiple energy conversion units are connected in parallel or cascaded, existing technologies typically achieve coordination through current sharing control, power droop, and unit rotation, but these methods assume that the energy path of each energy conversion unit is unidirectional and equivalent. However, this assumption does not hold true in systems where DC-DC, DC-IC, and ICDC coexist. This application prioritizes energy conversion units with high dynamic response adaptation indices to undertake rapid power regulation; energy conversion units with low dynamic response adaptation indices are locked in a slowly changing power range, thereby achieving synergistic efficiency at the system level, rather than improving the efficiency of individual energy conversion units.

[0108] Based on the corresponding adjustment task weights, each response adjustment unit is controlled to dynamically adjust the load side.

[0109] The dynamic response adjustment of the load end specifically includes:

[0110] Based on the corresponding dynamic response adaptation index and rated output power, the response adjustment period and power adjustment threshold are set for each response adjustment unit.

[0111] The method for setting the response adjustment period for any response adjustment unit is as follows: set a baseline adjustment period; assign values ​​to the length of the response adjustment period of each response adjustment unit based on the baseline adjustment period and the dynamic response adaptation index, and the length of the response adjustment period of any response adjustment unit is negatively correlated with the corresponding dynamic response adaptation index.

[0112] Those skilled in the art can set the length of the baseline adjustment period based on actual needs, such as 10ms; optionally, the length of the response adjustment period of any response adjustment unit can be assigned as follows: set the range of the response adjustment period length centered on 10ms, such as 10±5ms; within the range of values, assign the corresponding length according to the dynamic response adaptation index, and the larger the dynamic response adaptation index, the smaller the length of the response adjustment period, so as to make full use of the control capability of the response adjustment unit with a larger dynamic response adaptation index.

[0113] The method for setting the power regulation threshold for any response regulation unit is as follows: assign a value to the power regulation threshold of each response regulation unit based on the rated output power and the dynamic response adaptation index, and the power regulation threshold of any response regulation unit is positively correlated with the corresponding rated output power and dynamic response adaptation index.

[0114] Optionally, the power regulation threshold for any response regulation unit can be assigned as follows: A ratio regulation coefficient, for example, 10%, is set. For any response regulation unit, the power regulation threshold is the product of 10% of the rated output power and its dynamic response adaptation index. The regulation amplitude of each response regulation unit is limited according to the magnitude of the dynamic response adaptation index to prevent over-adjustment caused by severe oscillations.

[0115] An energy deviation threshold is set for each response adjustment unit. At the beginning of each response adjustment cycle, the energy deviation of the response adjustment unit is calculated for any response adjustment unit. If the absolute value of the energy deviation is greater than the corresponding energy deviation threshold, the energy deviation is suppressed by adjusting the output power of the response adjustment unit, and the adjustment amount of the output power is less than or equal to the corresponding power adjustment threshold.

[0116] Those skilled in the art can set the length of the energy deviation threshold based on actual needs, for example, 0.05 times the product of the rated output power and the response adjustment cycle length; adjust the output power of the response adjustment unit to suppress the energy deviation. For example, when the energy deviation is less than 0, the response adjustment unit provides excessive energy to the load, and the absolute value of the energy deviation is reduced by decreasing the output power, i.e., making the energy deviation approach 0, and the reduction in output power is not greater than the power adjustment threshold; when the energy deviation is greater than 0, the response adjustment unit provides insufficient energy to the load, and the energy deviation is made closer to 0 by increasing the output power, thereby ensuring stable power supply to the load. In practical applications, the output power of the response adjustment unit can be adjusted by adjusting the PWM duty cycle, phase offset angle, etc. of the power switch.

[0117] ICDC or DC-IC outputs often face complex loads, such as data centers and electric drive systems. These loads are not steady-state but exhibit microsecond-level pulsations and multi-scale energy requests. Traditional solutions address this by increasing buffer size and switching frequency, which easily introduces modulation errors and switching losses, ultimately reducing power supply efficiency. This application, by selecting a response regulation unit and limiting the corresponding regulation frequency and intensity, can reduce meaningless high-frequency modulation losses and provide compatibility with complex loads for efficient and flexible grid connection.

[0118] Optionally, if at least m% of the response adjustment units have an absolute value of energy deviation greater than the corresponding energy deviation threshold in the most recent n consecutive response adjustment cycles, then the adjustment capability of the current response adjustment unit is insufficient, and the proportion of response adjustment units is increased. For example, the top 30% of energy conversion units with the largest dynamic response adaptation index can be selected as response adjustment units. M and n are both positive integers.

[0119] The method for calculating the energy deviation of any response regulation unit is as follows: the total energy deviation at the load end is calculated based on the load voltage at the start of the response regulation cycle and at the start of the previous response regulation cycle; the energy deviation of the response regulation unit is the product of the total energy deviation and the corresponding regulation task weight.

[0120] Optionally, the total energy deviation at the load end is calculated using the capacitor energy storage formula; the equivalent capacitance at the load end is obtained; the difference between the square of the load voltage at the beginning of the previous response adjustment cycle and the square of the load voltage at the beginning of the current response adjustment cycle is calculated and multiplied by 0.5 times the equivalent capacitance to obtain the total energy deviation at the load end.

[0121] Example 2

[0122] This embodiment is the second embodiment of this application; it is based on the same inventive concept as Embodiment 1, and refers to... Figure 2 This embodiment introduces an energy storage multi-module collaborative energy conversion efficiency enhancement system, including a flexible grid-connected solid-state transformer and a control subsystem; the flexible grid-connected solid-state transformer includes an input module, an energy conversion module, and an output module; wherein, the input module includes multiple DC access units; any DC access unit is used to connect to the current output from an external DC source;

[0123] The external DC sources include, but are not limited to, photovoltaic DC sources, DC sources after wind power rectification, and the DC side of energy storage systems. These DC sources have different voltage levels and significantly different power fluctuation characteristics. In this embodiment, the DC sources are not directly connected to a unified DC bus, but are connected to their respective energy conversion units to achieve multi-source flexible grid connection.

[0124] The energy conversion module includes multiple energy conversion units; each energy conversion unit is connected to a DC input unit and is used to convert the current input to the DC input unit into energy.

[0125] The energy conversion unit includes, but is not limited to, a DC-DC conversion unit, a DC-IC conversion unit, and an ICDC interface unit. The energy conversion includes voltage level transformation and current amplitude adjustment to output stable DC power that meets the power supply requirements of the load.

[0126] The output module is used to integrate the current output by the energy conversion unit and supply power to the load.

[0127] The load end includes directly connected loads, such as DC loads in data centers and electric drive systems, and can also be a grid-connected interface. For example, the output module can combine the DC power output from multiple energy conversion units and convert it into AC power through DC-AC conversion to achieve flexible grid connection with the AC power grid. The energy demand of the load end is often not smooth and has a single time scale; it exhibits high-frequency fluctuations, periodic patterns, and sudden changes.

[0128] The control subsystem includes a data acquisition module, a data processing module, and a control module; wherein, the data acquisition module is used to acquire load change data at the load end and operating status data of each energy conversion unit;

[0129] The data processing module calculates the dynamic characteristic intensity of the energy at the load end based on the load change data, and calculates the dynamic response adaptation index of each energy conversion unit based on the operating status data.

[0130] The control module determines whether there is significant energy fluctuation at the load end based on the dynamic characteristic intensity; if so, the control module selects the response adjustment unit based on the dynamic response adaptation index and controls the response adjustment unit to perform dynamic response adjustment on the load end.

[0131] The specific functions of each module described above are implemented with reference to the relevant content in the energy storage multi-module collaborative energy conversion efficiency enhancement method described in Example 1, and will not be repeated here.

[0132] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of protection of this application, and these forms are all within the protection scope of this application.

Claims

1. A method for enhancing energy conversion efficiency through multi-module collaborative energy storage, characterized in that: Includes the following steps: The load change data at the load end is sampled at high frequency, and the operating status data of each energy conversion unit is collected; Based on the load change data, the energy change characteristics at the load end are analyzed on a time scale, and the intensity of dynamic characteristics characterizing rapid energy changes is calculated. Based on the intensity of the dynamic characteristics, determine whether there are significant energy fluctuations at the load end; If present, a dynamic response strategy is built for the load side, including: The dynamic response adaptation index of each energy conversion unit is calculated based on the aforementioned operating status data; The operating status data of any energy conversion unit includes input power and output power; Methods for calculating the dynamic response adaptation index of any energy conversion unit include: A power supply detection window is set; the power supply detection window contains the most recent N times; N is a positive integer; Extract the input power of the energy conversion unit at N time points within the power supply detection window, and calculate the standard deviation as the input index of the energy conversion unit; The first fit of the energy conversion unit is assigned a value based on the input index. Extract the output power of the energy conversion unit at N times within the power supply detection window, and calculate the average value as the output index of the energy conversion unit; Obtain the rated output power of the energy conversion unit, and assign a value based on the rated output power and the second fit of the output index of the energy conversion unit; The product of the first fit and the second fit is calculated and used as the dynamic response fit index of the energy conversion unit. Response adjustment units are selected based on the dynamic response adaptation index, and the adjustment task weight of each response adjustment unit is calculated. Based on the corresponding adjustment task weights, each response adjustment unit is controlled to dynamically adjust the load side; The dynamic response adjustment of the load end specifically includes: Based on the corresponding dynamic response adaptation index and rated output power, the response adjustment period and power adjustment threshold are set for each response adjustment unit. An energy deviation threshold is set for each response adjustment unit. At the beginning of each response adjustment cycle, the energy deviation of the response adjustment unit is calculated for any response adjustment unit. If the absolute value of the energy deviation is greater than the corresponding energy deviation threshold, the energy deviation is suppressed by adjusting the output power of the response adjustment unit, and the adjustment amount of the output power is less than or equal to the corresponding power adjustment threshold.

2. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 1, characterized in that: The load change data includes load voltage and load current; Based on the load change data, a time-scale analysis of the energy change characteristics at the load end is performed, specifically including: The instantaneous power at the load terminal at each moment is calculated based on the load voltage and load current, and the instantaneous power sequence at the load terminal is constructed. Set up a multi-scale analysis window; perform time-scale analysis based on the instantaneous power sequence and the multi-scale analysis window, and construct an energy change sequence; The intensity of dynamic features characterizing rapid energy changes is calculated based on the energy change sequence; Wherein, the multi-scale analysis window includes a first-scale window; the energy change sequence includes a first energy sequence; The method for constructing the first energy sequence includes: calculating the first scale power at each moment based on the instantaneous power sequence and the first scale window, and constructing the first energy sequence; the first energy sequence is a time series of the first scale power; Methods for calculating the first-scale power at any given time include: Mark any time point as the first target time point; extract the instantaneous power sequence through a first scale window, with the endpoint of the first scale window being the first target time point; calculate the mean of the instantaneous power at each time point within the first scale window as the first smoothed power; calculate the difference between the instantaneous power at the first target time point and the first smoothed power as the first scale power at the first target time point.

3. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 2, characterized in that: The multi-scale analysis window also includes a second-scale window, and the length of the first-scale window is less than the length of the second-scale window; The energy change sequence also includes a second energy sequence; The method for constructing the second energy sequence includes: calculating the second-scale power at each moment based on the instantaneous power sequence and the second-scale window, and constructing the second energy sequence; the second energy sequence is a time series of the second-scale power. Methods for calculating the second-scale power at any given time include: Mark any time point as the second target time point; extract the instantaneous power sequence through a second scale window, with the endpoint of the second scale window being the second target time point; calculate the mean of the instantaneous power at each time point within the second scale window as the second smoothed power; calculate the difference between the first smoothed power and the second smoothed power at the second target time point as the second scale power at the second target time point.

4. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 3, characterized in that: The dynamic feature intensity includes a first feature intensity and a second feature intensity; The method for calculating the first feature intensity includes: setting a load detection window, wherein the load detection window contains the most recent M time moments; M is a positive integer; Based on the load detection window, the instantaneous power of the most recent M moments is extracted from the instantaneous power sequence; the absolute value of the instantaneous power of the M moments within the load detection window is taken and the mean is calculated as a reference characteristic index; Based on the load detection window, the first-scale power of the most recent M moments is extracted from the first energy sequence; the absolute value of the first-scale power of the M moments within the load detection window is taken and the mean is calculated as the first feature index. Calculate the ratio of the first feature index to the reference feature index, and use it as the first feature intensity; The method for calculating the intensity of the second feature includes: Based on the load detection window, the second-scale power of the most recent M moments is extracted from the second energy sequence; the absolute value of the second-scale power of the M moments within the load detection window is taken and the mean is calculated as the second feature index. The ratio of the second feature index to the reference feature index is calculated as the second feature strength.

5. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 4, characterized in that: The methods for determining whether there are significant energy fluctuations at the load end include: Set a first intensity threshold and a second intensity threshold; if the first characteristic intensity is greater than the first intensity threshold or the second characteristic intensity is greater than the second intensity threshold, then there is a significant energy fluctuation at the load end.

6. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 5, characterized in that: The method for filtering the response adjustment unit is as follows: The energy conversion units are sorted in descending order of dynamic response adaptation index; The energy conversion unit that ranks first according to the preset selection ratio is selected as the response adjustment unit; The method for calculating the adjustment task weight of any response adjustment unit includes: assigning a value to the adjustment task weight of each response adjustment unit based on the dynamic response adaptation index; wherein, the adjustment task weight of any response adjustment unit is positively correlated with the corresponding dynamic response adaptation index, and the sum of the adjustment task weights of all response adjustment units is 1.

7. The energy storage multi-module collaborative energy conversion efficiency enhancement method as described in claim 6, characterized in that: The method for setting a response adjustment period for any response adjustment unit includes: setting a baseline adjustment period; assigning a value to the length of the response adjustment period of each response adjustment unit based on the baseline adjustment period and the dynamic response adaptation index, wherein the length of the response adjustment period of any response adjustment unit is negatively correlated with the corresponding dynamic response adaptation index. The method for setting a power regulation threshold for any response regulation unit includes: assigning a power regulation threshold for each response regulation unit based on the rated output power and the dynamic response adaptation index, wherein the power regulation threshold of any response regulation unit is positively correlated with the corresponding rated output power and the dynamic response adaptation index; The method for calculating the energy deviation of any response regulation unit is as follows: the total energy deviation at the load end is calculated based on the load voltage at the start of the response regulation cycle and at the start of the previous response regulation cycle; the energy deviation of the response regulation unit is the product of the total energy deviation and the corresponding regulation task weight.

8. A multi-module energy storage collaborative energy conversion efficiency enhancement system, used to implement the multi-module energy storage collaborative energy conversion efficiency enhancement method as described in any one of claims 1-7, characterized in that: Includes flexible grid-connected solid-state transformers and control subsystems; The flexible grid-connected solid-state transformer includes an input module, an energy conversion module, and an output module; wherein, the input module includes multiple DC access units; any DC access unit is used to connect to the current output from an external DC source; The energy conversion module includes multiple energy conversion units; each energy conversion unit is connected to a DC input unit and is used to convert the current input to the DC input unit into energy. The output module is used to integrate the current output by the energy conversion unit and supply power to the load. The control subsystem includes a data acquisition module, a data processing module, and a control module; wherein, the data acquisition module is used to acquire load change data at the load end and operating status data of each energy conversion unit; The data processing module calculates the dynamic characteristic intensity of the energy at the load end based on the load change data, and calculates the dynamic response adaptation index of each energy conversion unit based on the operating status data. The control module determines whether there is significant energy fluctuation at the load end based on the dynamic characteristic intensity; if so, the control module selects response adjustment units based on the dynamic response adaptation index and controls each response adjustment unit to perform dynamic response adjustment at the load end.

Images