Electric ship virtual power station dynamic aggregation method based on low-temperature working condition

Abstract

The invention discloses an electric ship virtual power station dynamic aggregation method based on a low-temperature working condition, and relates to the technical field of electric ship virtual power stations, and the method comprises the steps: obtaining the real-time temperature and charging and discharging efficiency of a battery pack of an electric ship in a low-temperature environment; the low-temperature performance coefficient of each electric ship is calculated, and electric ship clusters are divided according to the change trend and distribution characteristics of the low-temperature performance coefficient; calculating an initial polymerization degree in combination with the cluster performance level and the number of electric ships; calculating an updated polymerization degree according to the initial polymerization degree change trend; and a target aggregation electric ship cluster is screened through normalization processing and a preset threshold value, a charging and discharging strategy is dynamically adjusted based on a low-temperature performance coefficient, and a high-performance cluster is preferentially scheduled. According to the method, through multi-dimensional performance evaluation, dynamic cluster division and adaptive aggregation degree calculation, the resource scheduling efficiency and reliability of the virtual power station in the low-temperature environment are improved, and the method is suitable for efficient aggregation and power grid adjustment of the electric ship virtual power station in the low-temperature working condition.

Description

Technical Field

[0001] The present invention relates to a dynamic aggregation method of an electric ship virtual power station based on low-temperature working conditions, belonging to the technical field of electric ship virtual power stations. Background Art

[0002] As the global energy structure transitions toward a low-carbon future, electric ships, as a key vehicle for green shipping, are seeing increasing adoption. Virtual power plant technology aggregates dispersed electric ship resources, transforming them into a dispatchable energy aggregate. This can provide peak-shaving and frequency-regulating services to the power grid, enhancing renewable energy absorption capacity. However, in low-temperature environments, the battery performance of electric ships significantly degrades, manifesting as fluctuations in charge and discharge efficiency, capacity decay, and lifespan loss due to lower battery pack temperatures. This makes it difficult for traditional virtual power plant aggregation methods based on normal temperature conditions to effectively adapt to resource scheduling needs in low-temperature scenarios.

[0003] Currently, electric ship virtual power plants (VPPs) face the following major technical challenges in low-temperature operation: First, the impact of low temperatures on ship battery performance varies significantly from vessel to vessel. Factors such as battery type, age, and thermal management system efficiency lead to significant variations in real-time battery pack temperature and charge / discharge efficiency under the same low-temperature conditions. Adopting a unified aggregation strategy could result in underutilization of high-performing ship resources, while forcing poorer-performing ships into scheduling could lead to safety risks or inefficiencies.

[0004] Second, existing aggregation methods are mostly based on static parameter designs and lack the ability to respond to dynamic changes in ship performance under low-temperature conditions. Low-temperature environments are not constant. Ships may experience temperature gradients and start-stop operations during navigation, causing battery performance to fluctuate dynamically over time. Traditional methods are unable to capture these changes in real time and adjust aggregation strategies, affecting the regulation accuracy and reliability of the virtual power plant.

[0005] Third, cluster division and aggregation calculations fail to fully consider key factors influencing low-temperature performance. Existing technologies typically classify ship clusters based on a single dimension (such as battery capacity or charge / discharge power), without integrating low-temperature-sensitive indicators such as temperature adaptability and efficiency stability. Aggregation calculations also often overlook the coupled relationship between cluster size and performance concentration trends, resulting in aggregation results that fail to accurately reflect the actual aggregation value of ships in low-temperature environments.

[0006] Fourth, the charging and discharging strategy fails to dynamically match low-temperature performance. During grid regulation, if ship clusters with excellent low-temperature performance are not prioritized, this can lead to a decrease in overall regulation efficiency and even safety hazards such as battery thermal runaway caused by the frequent dispatch of ships with poor performance.

[0007] Furthermore, low temperatures slow the chemical reaction rate of batteries, increasing internal resistance and significantly increasing energy losses during charging and discharging, further complicating virtual power plant aggregation. Existing technologies lack a comprehensive dynamic aggregation method for electric ship virtual power plants tailored to low-temperature conditions. This approach lacks a comprehensive technical solution encompassing performance evaluation, cluster division, aggregation calculation, and scheduling strategies, making it difficult to meet the grid's demand for highly reliable and flexible energy aggregates. Summary of the Invention

[0008] The present invention aims to solve the problem of accurate clustering, dynamic aggregation and optimized scheduling of electric ship resources under low temperature conditions.

[0009] The technical solution of the present invention:

[0010] A dynamic aggregation method for an electric ship virtual power station based on low-temperature working conditions, the specific steps are as follows:

[0011] S1: Obtain the real-time temperature of the battery pack and the charging and discharging efficiency of the battery pack of each electric ship in a low-temperature environment;

[0012] S2: At any sampling moment, the low-temperature performance coefficient of each electric ship is obtained based on the real-time temperature of the battery pack of each electric ship and the distribution of the charge and discharge efficiency;

[0013] S3: Obtain multiple electric ship clusters based on the change trend and distribution characteristics of the low-temperature performance coefficients of all electric ships;

[0014] S4: obtaining the initial aggregation degree of each electric ship cluster at each sampling moment according to the low temperature performance coefficient level of the electric ships in each electric ship cluster and the number of electric ships in each electric ship cluster;

[0015] S5: Obtain updated aggregation degree according to the changing trend of the initial aggregation degree of each electric ship cluster at different sampling moments;

[0016] S6: Determine the target aggregated electric ship cluster according to the updated aggregation degree of each electric ship cluster.

[0017] Specifically, the method for obtaining the low-temperature performance coefficient in step S2 includes:

[0018] S11: At any sampling moment, the temperature adaptation degree of each electric vessel is obtained according to the change of the real-time temperature of the battery pack of each electric vessel relative to the lowest real-time temperature of the battery pack of other electric vessels;

[0019] S12: Obtaining a first difference between the charging and discharging efficiency of each electric vessel and the average charging and discharging efficiency;

[0020] S13: normalizing the first difference to obtain a normalized difference;

[0021] S14: Calculating the preset even power of the standardized differences between all charging and discharging efficiencies of each electric vessel and the average charging and discharging efficiency, and then finding the average to obtain the efficiency fluctuation degree;

[0022] S15: Obtaining a low-temperature performance coefficient of each electric vessel according to the temperature adaptability and efficiency fluctuation of each electric vessel, wherein the temperature adaptability and efficiency fluctuation are positively correlated with the low-temperature performance coefficient.

[0023] Specifically, the method for obtaining the electric ship cluster in step S3 includes:

[0024] S31: Filtering multiple reference low-temperature performance coefficients based on the change trend and distribution characteristics of the low-temperature performance coefficients of all electric ships;

[0025] S32: Sequentially obtain ships corresponding to the low-temperature performance coefficients within the range of every two reference low-temperature performance coefficients to form an electric ship cluster.

[0026] Specifically, the method for screening the reference low-temperature performance coefficient in step S31 includes:

[0027] S311: Construct a descending sequence of low-temperature performance coefficients for all electric ships;

[0028] S312: Obtaining a low-temperature performance coefficient difference sequence of the low-temperature performance coefficient descending sequence, and obtaining multiple maximum value points of the low-temperature performance coefficient difference sequence;

[0029] S313: The low-temperature coefficient of performance corresponding to the maximum point, the maximum low-temperature coefficient of performance, and the minimum low-temperature coefficient of performance are used as reference low-temperature coefficients of performance.

[0030] Specifically, the method for obtaining the initial degree of polymerization in step S4 includes:

[0031] S41: obtaining an average value of the low-temperature performance coefficients of all electric ships in each electric ship cluster as the overall performance level;

[0032] S42: Sort the electric ship clusters from largest to smallest according to the overall performance levels, and obtain the performance concentration level of each electric ship cluster according to the difference characteristics between the overall performance level of each electric ship cluster and the adjacent overall performance levels and the central trend of the overall performance levels of all electric ship clusters, wherein the difference characteristics are positively correlated with the performance concentration level, and the central trend is negatively correlated with the performance concentration level;

[0033] S43: Calculate the ratio of the number of electric ships in each electric ship cluster to the number of all electric ships as an aggregation factor;

[0034] S44: Fusing the performance concentration degree and aggregation factor of each electric ship cluster to obtain the initial aggregation degree of each ship cluster at each sampling moment.

[0035] Specifically, the method for obtaining the maximum point in step S42 is:

[0036] The AMPD algorithm is used to obtain the maximum point of the low-temperature performance coefficient difference series.

[0037] Specifically, the method for obtaining the updated aggregation degree in step S5 includes:

[0038] S51: Obtain the mean of the initial aggregation degree of the electric ship cluster at all sampling moments as the overall aggregation level;

[0039] S52: Obtain an updated aggregation degree of each electric ship cluster according to the overall aggregation level of each electric ship cluster and the change in the initial aggregation degree between adjacent sampling moments.

[0040] Specifically, step S52 includes:

[0041] S521: Selecting the adjacent sampling moments corresponding to the changes that are not equal to the preset changes as reference adjacent sampling moments;

[0042] S522: Obtain a first ratio between the overall aggregation level and the amount of change in the initial aggregation degree between each reference adjacent sampling moment, and calculate the cumulative sum of the overall aggregation level and the first ratios between all reference adjacent sampling moments as the updated aggregation degree of each ship cluster.

[0043] Specifically, the method for obtaining the target aggregated ship cluster in step S6 is:

[0044] The updated aggregation degree of each ship cluster is normalized. If the normalized result is greater than or equal to the preset aggregation threshold, the corresponding ship cluster is determined to be the target aggregation ship cluster.

[0045] Specifically, the method further includes:

[0046] The charging and discharging strategies of the virtual power station are dynamically adjusted according to the low-temperature performance coefficient of the target aggregated ship cluster, and the ship cluster with a high low-temperature performance coefficient is preferentially dispatched to participate in grid regulation.

[0047] Beneficial effects of the present invention:

[0048] 1. In terms of low-temperature performance assessment, a low-temperature performance coefficient is constructed by integrating temperature adaptability and efficiency fluctuation, achieving accurate quantification of the comprehensive performance of electric ships in low-temperature environments. Temperature adaptability reflects the deviation of the electric ship's battery pack temperature from the minimum temperature, effectively screening electric ships with temperature advantages in low-temperature environments. Efficiency fluctuation is calculated by taking the even-power mean of the standardized difference, highlighting the stability of charge and discharge efficiency and avoiding the misclassification of electric ships as high-performance ships due to drastic efficiency fluctuations caused by low temperatures. Compared with traditional single-parameter evaluation methods, this dual-index evaluation system more comprehensively captures the dual impact of low temperatures on battery performance: temperature sensitivity and efficiency stability, providing a reliable basis for subsequent clustering.

[0049] 2. In terms of electric ship clustering, by screening the reference low-temperature performance coefficient based on the descending sequence of the low-temperature performance coefficient and the maximum point of the differential sequence, the key nodes of low-temperature performance changes can be dynamically identified, and electric ships can be divided into clusters with significant performance differences. Compared with the traditional division method based on fixed thresholds or simple clustering algorithms, this method can adapt to the distribution characteristics of low-temperature performance, ensure that the electric ships in each cluster have highly similar low-temperature performance, reduce the performance dispersion within the cluster, and thus reduce scheduling conflicts and efficiency losses caused by internal differences during the aggregation process. For example, by accurately locating the maximum point of the differential sequence through the AMPD algorithm, cluster division deviations caused by noisy data can be avoided, and the robustness of cluster division can be improved.

[0050] 3. The calculation of the initial aggregation degree integrates the overall performance level, performance concentration, and aggregation factor, taking into account the performance advantages, performance consistency, and scale ratio of the cluster. The mean calculation of the overall performance level provides a basic performance evaluation of the cluster; the performance concentration degree quantifies the uniqueness of the cluster in the performance ranking through a dual analysis of the performance differences between adjacent clusters and the overall concentration trend; the aggregation factor directly reflects the scale weight of the cluster in the entire electric ship group. This multi-factor fusion calculation method enables the initial aggregation degree to comprehensively measure the aggregation potential of the cluster, avoiding the one-sidedness of relying solely on a single dimension of performance or scale. For example, a cluster with medium performance but large scale may obtain a higher initial aggregation degree due to a higher aggregation factor, enabling it to compete with high-performance small clusters during the aggregation process, which is more in line with the scheduling needs of the virtual power plant for the total amount of resources.

[0051] 4. The calculation of the updated aggregation degree incorporates dynamic analysis over time. By cumulatively analyzing the overall aggregation level and the changes in adjacent moments, we capture the changing trends in a cluster's aggregation capabilities. The overall aggregation level, as a long-term average indicator, ensures the stability of the aggregation degree; while the reference to the changes in adjacent moments and the accumulation of the first ratio highlight recent fluctuations in aggregation capabilities. This design allows the updated aggregation degree to reflect both the historical performance of the cluster and respond to real-time changes. For example, a cluster with a moderate long-term aggregation level but a significant recent increase in aggregation capabilities can be re-identified as a high-value aggregation target through dynamic adjustments to the updated aggregation degree, avoiding the delayed response to emerging high-quality resources that often occurs with traditional static aggregation methods.

[0052] 5. Targeted aggregation of electric vessel clusters is determined by combining normalization and preset aggregation thresholds, enabling standardized screening of cluster aggregation qualifications. Normalization eliminates dimensional differences in the updated aggregation degrees of different clusters, making them comparable. Dynamically adjustable preset aggregation thresholds allow for flexible adjustment of aggregation scale based on grid demand, expanding the aggregation range to increase regulation capacity during peak loads and narrowing the range to improve regulation quality during low loads. Furthermore, a charging and discharging strategy based on low-temperature performance coefficients prioritizes high-performance clusters, fully leveraging the efficiency of high-quality resources, reducing the ineffective deployment of low-performance vessels, minimizing battery loss and safety risks, and improving the response speed and accuracy of grid regulation.

[0053] Through the low-temperature adaptability design of the entire process, this invention constructs a closed-loop technical solution from performance evaluation to scheduling strategy, effectively solving the key technical difficulties of virtual power station aggregation of electric ships under low-temperature conditions, and providing important technical support for improving the coordinated development level of green shipping and smart grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a workflow diagram of the dynamic aggregation method of the electric ship virtual power station based on low temperature conditions;

[0055] Figure 2 A design diagram for a method of obtaining a low temperature coefficient of performance;

[0056] Figure 3 A design diagram for the electric ship cluster division method;

[0057] Figure 4 This is the design diagram of the method for calculating the initial degree of polymerization. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] like Figure 1-Figure 4 , the present invention is specifically implemented as follows:

[0060] S1: Obtain the real-time temperature of the battery pack and the charging and discharging efficiency of the battery pack of each electric ship in a low-temperature environment;

[0061] S2: At any sampling moment, the low-temperature performance coefficient of each electric ship is obtained based on the real-time temperature of the battery pack of each electric ship and the distribution of the charge and discharge efficiency;

[0062] S3: Obtain multiple electric ship clusters based on the change trend and distribution characteristics of the low-temperature performance coefficients of all electric ships;

[0063] S4: obtaining the initial aggregation degree of each electric ship cluster at each sampling moment according to the low temperature performance coefficient level of the electric ships in each electric ship cluster and the number of electric ships in each electric ship cluster;

[0064] S5: Obtain updated aggregation degree according to the changing trend of the initial aggregation degree of each electric ship cluster at different sampling moments;

[0065] S6: Determine the target aggregated electric ship cluster according to the updated aggregation degree of each electric ship cluster.

[0066] The specific steps for obtaining the low-temperature performance coefficient in step S2 are:

[0067] S21: At any sampling moment, the temperature adaptability of each electric vessel is determined based on the change in the real-time battery pack temperature of each electric vessel relative to the lowest real-time battery pack temperature among all other electric vessels. The real-time battery pack temperature of each electric vessel is compared with the lowest real-time battery pack temperature among all other electric vessels. The difference between the two is first calculated. This difference intuitively reflects the degree to which the current battery pack temperature of the electric vessel deviates from the lowest temperature. Subsequently, to ensure uniformity and comparability between temperature comparisons across different electric vessels, this difference is normalized. This normalization process maps the difference to a specific interval, such as [0,1], based on the distribution range of the battery pack temperatures across the entire fleet of electric vessels. This results in the temperature adaptability of each electric vessel. This temperature adaptability value represents the proximity of the electric vessel's battery pack temperature to the optimal low-temperature state in the current low-temperature environment. A larger value indicates that the battery pack temperature of the electric vessel is closer to the lowest temperature within the entire fleet of electric vessels, indicating a greater temperature adaptability in low-temperature environments.

[0068] S22: Calculate a first difference between the charge and discharge efficiency of each electric vessel and the average charge and discharge efficiency. The average charge and discharge efficiency is calculated by statistically averaging the charge and discharge efficiencies of all electric vessels at the sampling time. This first difference can initially reflect the difference between the charge and discharge efficiency of a single electric vessel and the overall average level.

[0069] S23: Normalize the first difference to obtain a standardized difference. Because the charge and discharge efficiency ranges of different electric vessels may differ, the first difference must be normalized to eliminate the impact of these differences in dimension and range. The normalization process utilizes specific mathematical methods, such as a standardization formula based on standard deviation and mean, to convert the first difference into a standardized difference with the same scale and distribution characteristics.

[0070] S24: After calculating the preset even power of the standardized difference between all the charge and discharge efficiencies and the average charge and discharge efficiency of each electric vessel, the average is calculated to obtain the degree of efficiency fluctuation. For each electric vessel, after obtaining a series of standardized differences, the preset even power, such as the fourth power, of the standardized difference between all its charge and discharge efficiencies and the average charge and discharge efficiency is calculated. The preset even power is chosen because it can better highlight the impact of standardized differences with larger differences on the results, while avoiding the positive and negative offset problems that may be caused by odd powers. The average operation is performed on these calculated fourth power results, and the final average obtained is the degree of efficiency fluctuation of the electric vessel. The degree of efficiency fluctuation reflects the stability of the efficiency of the electric vessel during the charging and discharging process. The smaller the value, the more stable the charging and discharging efficiency of the electric vessel and the less affected by the low temperature environment.

[0071] S25: Based on the temperature adaptability and efficiency fluctuation of each electric vessel, the low-temperature performance coefficient of each electric vessel is obtained, where the temperature adaptability and efficiency fluctuation are positively correlated with the low-temperature performance coefficient. After obtaining the temperature adaptability and efficiency fluctuation of each electric vessel, these two key indicators need to be integrated to obtain the low-temperature performance coefficient of each electric vessel. The integration method adopts weighted summation, that is, the weights of the temperature adaptability and efficiency fluctuation are set according to actual needs. The setting of these weights needs to comprehensively consider the impact of temperature and charge and discharge efficiency on the performance of the electric vessel in low-temperature environments. In practical applications, reasonable weight values ​​can be determined by analyzing a large amount of historical data and combining expert experience. Moreover, the temperature adaptability and efficiency fluctuation are positively correlated with the low-temperature performance coefficient. That is, the higher the temperature adaptability and the lower the efficiency fluctuation, the larger the calculated low-temperature performance coefficient, indicating that the overall performance of the electric vessel in low-temperature conditions is better.

[0072] Through this rigorous and complex series of calculation steps, the performance of each electric ship in low-temperature environments can be comprehensively and accurately evaluated, providing a solid and reliable data foundation for subsequent key operations such as electric ship cluster division, aggregation degree calculation, and target aggregation electric ship cluster determination based on the low-temperature performance coefficient. Each calculation link is closely linked, and the results of the previous step directly affect the accuracy of the subsequent steps. Deviations in any data processing process may lead to inaccurate final low-temperature performance coefficient assessment, thereby affecting the effectiveness and reliability of the entire electric ship virtual power plant dynamic aggregation method.

[0073] Step S3: Based on the changing trend and distribution characteristics of the low-temperature performance coefficients of all electric ships, multiple electric ship clusters are obtained. The specific steps are:

[0074] S31: In the process of dividing electric ship clusters, the first step is to conduct a systematic analysis of the low-temperature performance coefficients of all electric ships. By exploring their changing trends and distribution characteristics, a number of representative reference low-temperature performance coefficients are screened out. These coefficients will serve as the key basis for dividing electric ship clusters.

[0075] S311: First, construct a descending sequence of the low-temperature performance coefficients of all electric ships. In this process, the low-temperature performance coefficients calculated for each electric ship need to be sorted and arranged in descending order. This sorting method can clearly show the advantages and disadvantages of different electric ships in terms of low-temperature performance, providing an intuitive data basis for subsequent analysis. For example, if there are 100 electric ships, these 100 low-temperature performance coefficients are arranged in descending order, so that the coefficient of the electric ship with the best performance is at the front of the sequence, and the coefficient of the electric ship with the worst performance is at the end of the sequence.

[0076] S312: After constructing the descending sequence, the next step is to calculate the low-temperature performance coefficient difference sequence for that sequence. This difference sequence is calculated by sequentially calculating the difference between two adjacent low-temperature performance coefficients. This step highlights the magnitude of change between adjacent electric vessel low-temperature performance coefficients. By analyzing these magnitudes, we can more accurately understand the changing trend of the low-temperature performance coefficient for the entire electric vessel fleet. For example, in a descending sequence, the difference between the first and second coefficients, the difference between the second and third coefficients, and so on, is calculated until the differences of all adjacent coefficients in the entire sequence are calculated, resulting in a complete low-temperature performance coefficient difference sequence. After obtaining the difference sequence, the AMPD (Adaptive Multi-Parameter Decomposition) algorithm is used to obtain multiple maximum points in the sequence. The AMPD algorithm is an algorithm that effectively processes complex signals and extracts key signal features. In this scenario, the AMPD algorithm analyzes and processes the low-temperature performance coefficient difference sequence to identify the points with the largest magnitude of change in the sequence, known as the maximum points. These maximum points represent key turning points in the low-temperature performance coefficient change process, indicating significant changes in the low-temperature performance of electric vessels within certain ranges. For example, in a differential sequence, there may be certain positions where the differences between the adjacent positions increase significantly. After being processed by the AMPD algorithm, the points corresponding to these positions will be identified as maximum points.

[0077] S313: After determining the maximum point, the corresponding low-temperature performance coefficient (COP) is used as the reference COP, along with the maximum and minimum COPs in the entire sequence. The coefficients corresponding to the maximum points are selected because they represent key points in the COP's variation; the maximum and minimum COPs define the range of the COP for the entire electric ship cluster. These coefficients together form the reference COP set, providing clear demarcation criteria for the electric ship cluster.

[0078] S32: After obtaining the reference low-temperature performance coefficient, the construction of the electric ship cluster begins. The specific operation is to obtain the electric ships corresponding to the low-temperature performance coefficient within each two reference low-temperature performance coefficients in turn. That is to say, with each two reference low-temperature performance coefficients as the upper and lower limits of the interval, all electric ships whose low-temperature performance coefficients fall within the interval are selected, and these electric ships are grouped together to form an electric ship cluster. For example, if there are reference low-temperature performance coefficients A, B, and C, then A and B, B and C will be used as interval boundaries to screen out electric ships that fall within the corresponding intervals, forming different electric ship clusters. In this way, electric ships with similar low-temperature performance coefficients can be divided into the same cluster, so that the electric ships in each cluster have a high similarity in low-temperature performance, which lays a solid foundation for the subsequent calculation of the aggregation degree of the electric ship cluster and the judgment of the target aggregated electric ship cluster. Each electric ship cluster represents a collection of electric ships with certain commonalities in low-temperature performance, which helps to classify and manage electric ships and optimize scheduling, thereby realizing efficient dynamic aggregation operation of electric ship virtual power stations under low-temperature conditions.

[0079] Step S4 obtains the initial aggregation degree of each electric ship cluster at each sampling moment according to the low temperature performance coefficient level of the electric ships in each electric ship cluster and the number of electric ships in each electric ship cluster. The specific steps are:

[0080] S41: When calculating the initial aggregation degree, each electric ship cluster must be processed carefully first. For each electric ship cluster, the low-temperature performance coefficients of all electric ships in the cluster are summarized, and then the overall performance level of the cluster is obtained by averaging. The overall performance level reflects the comprehensive performance of the electric ship cluster under low-temperature conditions and is a basic indicator for measuring the quality of the cluster. For example, there are 10 electric ships in a certain electric ship cluster, and their low-temperature performance coefficients are , then the overall performance level of the cluster The calculation formula is:

[0081]

[0082] in Represents the overall performance level, ( ) represents the low-temperature performance coefficient of each ship in the cluster.

[0083] S42: After obtaining the overall performance levels of all ship clusters, the electric ship clusters are ranked in descending order based on their overall performance levels. This step makes the performance relationships between the electric ship clusters clearer and more intuitive, facilitating further analysis. After the ranking is complete, an in-depth analysis begins of the differences between the overall performance level of each electric ship cluster and the adjacent overall performance levels. Specifically, the absolute difference in performance between adjacent clusters can be calculated by subtracting the two overall performance levels, or the ratio can be calculated to reflect the degree of performance difference between adjacent clusters through a relative proportional relationship. These differences can reflect the degree of dispersion in the performance of different electric ship clusters. The greater the difference, the more significant the performance gap between adjacent clusters. At the same time, the central tendency of the overall performance levels of all electric ship clusters must be calculated. Central tendency is an important statistic that describes the degree of concentration in data distribution. Common calculation methods include calculating variance, standard deviation, and other indicators. Taking variance as an example, it measures the degree of dispersion in the data by calculating the mean of the sum of the squares of the deviations between the overall performance level of each cluster and the average overall performance level of all clusters. Smaller variances indicate a more concentrated overall performance level across all electric ship clusters, with smaller differences between them. Conversely, larger variances indicate a more dispersed overall performance level across all electric ship clusters. By analyzing the differential characteristics and central tendency, and utilizing a specific computational model, the degree of performance concentration within each electric ship cluster can be determined. The degree of performance concentration here is positively correlated with the differential characteristics, meaning that the greater the performance differences between adjacent clusters, the higher the performance concentration within that cluster. It is negatively correlated with the central tendency, meaning that the more concentrated the overall performance level across all clusters, the lower the performance concentration within a single cluster.

[0084] S43: Calculate the ratio of the number of electric ships in each electric ship cluster to the number of all electric ships, and use this ratio as the aggregation factor. The aggregation factor reflects the proportion of the size of a single electric ship cluster in the entire electric ship cluster. For example, if the entire electric ship cluster has 100 ships and a certain electric ship cluster has 20 ships, then the aggregation factor of the electric ship cluster is , indicating that this cluster accounts for 20% of the total number of electric ships. A larger aggregation factor indicates that the electric ship cluster accounts for a larger proportion of the entire electric ship population, and is likely to have a greater influence in the subsequent aggregation process.

[0085] Finally, the performance concentration degree and aggregation factor of each electric ship cluster are fused to obtain the initial aggregation degree of each electric ship cluster at each sampling moment. The fusion method usually adopts the weighted summation method, that is, according to the actual needs and the importance of performance concentration degree and aggregation factor, appropriate weights are given to each of them. For example, if the weight given to performance concentration degree is , the weight assigned to the aggregation factor is ,and , the performance concentration of a certain electric ship cluster is , the aggregation factor is , then the initial aggregation degree of the electric ship cluster is The calculation formula is .in, represents the initial degree of polymerization, Represents the degree of performance concentration, represents the aggregation factor, and Respectively represent the weights of the corresponding indicators. Through such fusion calculations, the performance advantages and scale proportions of the electric ship cluster are comprehensively considered, so that the initial aggregation degree can fully and accurately reflect the potential and value of each electric ship cluster in participating in the virtual power station aggregation at the current sampling moment, and provide a key quantitative basis for further calculation and update of the aggregation degree based on the initial aggregation degree and judgment of the target aggregated electric ship cluster. The initial aggregation degree of each electric ship cluster is not fixed, but will be updated in real time as the sampling moment changes, according to the dynamic changes of factors such as the low-temperature performance coefficient of ships in the electric ship cluster and the number of electric ships, thereby ensuring that the entire electric ship virtual power station dynamic aggregation method can adapt to the ever-changing low-temperature working conditions and ship operating status.

[0086] Step S5: According to the changing trend of the initial aggregation degree of each electric ship cluster at different sampling times, the specific steps of obtaining the updated aggregation degree are as follows:

[0087] S51: First, determine the average of the initial aggregation degree of the electric ship cluster at all sampling moments, and use it as the overall aggregation level. The overall aggregation level reflects the average ability of the electric ship cluster to participate in aggregation over a period of time and is a basic indicator for measuring its aggregation stability. Taking a certain electric ship cluster as an example, assuming that at 5 consecutive sampling moments (denoted as to ), their initial polymerization degrees are , then the overall aggregation level is the average of these five values, that is, This value represents the average aggregation capacity of the cluster during this period and provides a benchmark reference for subsequent analysis.

[0088] S52: Analyze the change in the initial aggregation degree of each electric ship cluster between adjacent sampling moments. The change is calculated by subtracting the initial aggregation degree at the previous sampling moment from the initial aggregation degree at the next sampling moment. For example, for the above electric ship cluster, and The change between , and The change between , and The change between , and The change between These changes reflect the fluctuations in the cluster aggregation capacity at adjacent moments, which may be caused by factors such as changes in the battery temperature of electric ships, fluctuations in charging and discharging efficiency, or adjustments to the number of electric ships in the cluster.

[0089] After obtaining the changes of all adjacent sampling moments, the corresponding adjacent sampling moments that are not equal to the preset changes are screened out as reference adjacent sampling moments. The preset change is usually set according to the system's sensitivity to fluctuations in the degree of aggregation. For example, it can be set to 0 or a small value close to 0 (such as ±0.01) to filter out changes that are meaningless due to small fluctuations in data or measurement errors. Continuing with the above-mentioned electric ship cluster as an example, if the preset change is 0, then the adjacent moments with changes of -0.1, 0.2, and -0.1 ( 、 、 ) are all referenced to adjacent moments, and The change is -0.05. If its absolute value is less than the preset change (such as 0.01), it will not be included in the reference range.

[0090] For each reference adjacent sampling moment, the first ratio between the overall aggregation level and the change in the initial aggregation degree at that moment needs to be calculated. The calculation logic of the first ratio is to measure the impact of the change on the cluster aggregation capability trend through the relative relationship between the overall aggregation level and the change. For example, At this moment, the overall aggregation level is 0.59, the change is 0.2, and the first ratio can be understood as and The specific calculation method needs to be determined according to the system design rules. Assume that the ratio calculation is used here, that is, the first ratio is , the larger the value, the more significant the positive change is in improving the cluster aggregation ability when the aggregation level is high; on the contrary, if the change is negative (such as -0.1 at the moment), the first ratio may be negative, reflecting the inverse relationship between the aggregation level and the change trend.

[0091] S52: After completing the calculation of the first ratios of all reference adjacent moments, these ratios need to be accumulated and summed. The accumulated sum is the updated aggregation degree of each electric ship cluster. The accumulation and summation process is intended to comprehensively consider the trend of cluster aggregation capabilities at multiple effective fluctuation moments to avoid excessive impact of abnormal fluctuations at a single moment on the results. Continuing with the above example, assuming that the first ratios of the three reference adjacent moments are (correspond )、 (correspond )、 (correspond , ), the cumulative sum is , which is the updated aggregation degree of the electric ship cluster.

[0092] The physical significance of updating the aggregation degree lies in dynamically assessing the actual value of a cluster's current participation in virtual power plant aggregation by combining its long-term average aggregation capacity (overall aggregation level) with its short-term fluctuation trend (variation). If the updated aggregation degree is positive and large, it indicates that the cluster has maintained a high average aggregation level while its aggregation capacity has recently increased, making it suitable for key aggregation targets. If it is negative or small, it may indicate that the cluster's aggregation capacity is unstable or declining, requiring further observation or adjustment of the scheduling strategy.

[0093] In practice, the calculation process for the updated aggregation degree of different electric ship clusters is similar, but the specific values ​​will vary depending on the initial aggregation degree sequence, overall aggregation level, and distribution of changes in each cluster. For example, another ship cluster may have a small initial aggregation degree fluctuation at consecutive sampling moments, with most changes close to the preset value. This results in fewer reference adjacent moments and a smaller cumulative sum, and its updated aggregation degree may be closer to the overall aggregation level itself. This calculation method can flexibly adapt to the dynamic characteristics of different clusters, ensuring that the updated aggregation degree accurately reflects the real-time changes in their aggregation capacity, providing a reliable basis for subsequent judgment of the target aggregation electric ship cluster.

[0094] It's important to note that the preset change amount should be determined based on the actual system requirements. If it's set too strictly (e.g., the preset value is too small), it may result in too few adjacent reference moments, failing to effectively capture the cluster's true fluctuations. If it's set too loosely (e.g., the preset value is too large), it may include too many invalid changes, affecting the accuracy of the calculation results. Therefore, during system deployment, the preset change amount should be reasonably determined through statistical analysis of historical data or expert experience to ensure the scientific and effective calculation of the updated aggregation degree.

[0095] Through the above steps, the updated aggregation degree of each electric ship cluster can comprehensively reflect the changes in its aggregation ability in the time dimension, which not only avoids the one-sidedness of relying only on data at a single moment, but also overcomes the shortcomings of only considering the average level and ignoring dynamic trends. It provides more comprehensive and accurate quantitative support for the dynamic aggregation of electric ship virtual power stations under low-temperature conditions.

[0096] Step S6 determines the target aggregated electric ship cluster according to the updated aggregation degree of each electric ship cluster. The specific steps are:

[0097] When determining the target electric ship cluster, the update aggregation degree of each electric ship cluster must first be normalized. The purpose of normalization is to convert the update aggregation degrees of different electric ship clusters into a unified numerical range, facilitating comparison and judgment. For example, consider a scenario consisting of three ship clusters, A, B, and C, with update aggregation degrees of 12, 8, and 4, respectively. During normalization, these values ​​are typically mapped to a range, such as [0, 1], based on the maximum and minimum update aggregation degrees of all clusters. In this example, the maximum is 12 and the minimum is 4. For cluster A, the normalized value is calculated by first calculating the difference between its update aggregation degree and the minimum (12 - 4 = 8), then dividing it by the difference between the maximum and minimum values ​​(12 - 4 = 8), resulting in a normalized value of 1. Similarly, the normalized value for cluster B is (8 - 4) ÷ (12 - 4) = 0.5; and the normalized value for cluster C is (4 - 4) ÷ (12 - 4) = 0. Through such normalization operations, the update aggregation degrees that originally had large numerical differences are converted into comparable values, which facilitates subsequent judgment.

[0098] After the normalization process is completed, the normalized results of each electric ship cluster need to be compared with the preset aggregation threshold. The preset aggregation threshold is a key value determined based on multiple factors such as the operating requirements of the virtual power station, the grid regulation capabilities, and the overall performance of the ship cluster. It is used to screen out electric ship clusters suitable for aggregation. Assuming that the preset aggregation threshold is set to 0.6, for the three electric ship clusters mentioned above, the normalized result 1 of cluster A is greater than 0.6, so cluster A will be judged as the target aggregated electric ship cluster; while the normalized result 0.5 of cluster B is less than 0.6, and the normalized result 0 of cluster C is also less than 0.6, so clusters B and cluster C are not identified as target aggregated electric ship clusters in this judgment.

[0099] After determining the target aggregated electric ship cluster, it is necessary to dynamically adjust the charge and discharge strategy of the virtual power plant based on the low-temperature performance coefficient of the target aggregated electric ship cluster. Taking a virtual power plant containing multiple target aggregated electric ship clusters as an example, assume that the currently determined target aggregated electric ship clusters are D, E, and F. The average low-temperature performance coefficient of cluster D is 0.8, the average low-temperature performance coefficient of cluster E is 0.7, and the average low-temperature performance coefficient of cluster F is 0.6. Since the low-temperature performance coefficient is positively correlated with the performance of electric ships in low-temperature environments, the higher the low-temperature performance coefficient, the better the efficiency and stability of the electric ship during the charging and discharging process, and the more suitable it is for participating in grid regulation. Therefore, when adjusting the charge and discharge strategy, cluster D, which has the highest low-temperature performance coefficient, is prioritized for grid regulation, followed by cluster E, and finally cluster F.

[0100] In actual dispatching, prioritizing clusters of electric vessels with high low-temperature performance coefficients for grid regulation is crucial. For example, when the grid experiences a power shortage and requires additional energy, the electric vessels in cluster D are prioritized for discharge operations. Because these electric vessels perform exceptionally well in low-temperature environments, they can release energy from their batteries to the grid with high charge and discharge efficiency. Furthermore, the battery packs are less affected by low temperatures during discharge, maintaining a stable operating state and thus more efficiently meeting the grid's power needs. Furthermore, when the grid experiences excess power and requires energy storage, the vessels in cluster D are also prioritized for charging. These electric vessels can quickly and stably absorb energy, reducing charging time and energy loss.

[0101] For electric ship clusters that are not prioritized, such as cluster E and cluster F, they do not completely not participate in grid regulation. When the grid regulation demand is relatively low, or the prioritized electric ship clusters cannot meet all the regulation needs, these clusters will also participate in regulation in order of low-temperature performance coefficient. For example, when the electric ships in cluster D have reached their charge and discharge limits, and the grid still has regulation needs, the ships in cluster E will be started to participate in regulation. At the same time, during the scheduling process, the status of each electric ship cluster will be continuously monitored, including information such as the real-time temperature of the battery pack, charge and discharge efficiency, and remaining power. Once the status of an electric ship cluster is found to have changed, such as temperature increase leading to performance degradation, or the power reaching a critical value, the scheduling strategy will be adjusted in a timely manner to re-evaluate the participation order and regulation amount of each cluster.

[0102] Furthermore, the preset aggregation threshold is not fixed but needs to be dynamically adjusted based on factors such as the operating status of the virtual power plant, the real-time needs of the power grid, and changes in the overall performance of the electric vessel. For example, during peak grid load periods, when the demand for power regulation increases, the preset aggregation threshold can be appropriately lowered to allow more electric vessel clusters to be included in the target aggregation range to meet the grid's regulation needs. During low grid load periods, to ensure the quality and efficiency of regulation, the preset aggregation threshold can be appropriately raised, selecting only electric vessel clusters with better performance for regulation. Through such dynamic adjustment and refined scheduling, the aggregation advantages of the electric vessel virtual power plant in low-temperature conditions can be fully utilized, achieving efficient regulation and stable operation of the grid.

Claims

1. A dynamic aggregation method for electric ship virtual power station based on low temperature working conditions, characterized in that: The specific steps are as follows: S1: Obtain the real-time temperature of the battery pack and the charging and discharging efficiency of the battery pack of each electric ship in a low-temperature environment; S2: At any sampling moment, the low-temperature performance coefficient of each electric ship is obtained based on the real-time temperature of the battery pack of each electric ship and the distribution of the battery pack charge and discharge efficiency; S3: Obtain multiple electric ship clusters based on the change trend and distribution characteristics of the low-temperature performance coefficients of all electric ships; S4: obtaining the initial aggregation degree of each electric ship cluster at each sampling moment according to the low temperature performance coefficient level of the electric ships in each electric ship cluster and the number of electric ships in each electric ship cluster; S5: Obtain updated aggregation degree according to the changing trend of the initial aggregation degree of each electric ship cluster at different sampling moments; S6: Determine the target aggregated electric ship cluster according to the updated aggregation degree of each electric ship cluster.

2. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method for obtaining the low-temperature performance coefficient in step S2 includes: S11: At any sampling moment, the temperature adaptation degree of each electric vessel is obtained according to the change of the real-time temperature of the battery pack of each electric vessel relative to the lowest real-time temperature of the battery pack of other electric vessels; S12: Obtaining a first difference between the charging and discharging efficiency of each electric vessel and the average charging and discharging efficiency; S13: normalizing the first difference to obtain a normalized difference; S14: Calculating the preset even power of the standardized differences between all charging and discharging efficiencies of each electric vessel and the average charging and discharging efficiency, and then finding the average to obtain the efficiency fluctuation degree; S15: Obtaining a low-temperature performance coefficient of each electric vessel according to the temperature adaptability and efficiency fluctuation of each electric vessel, wherein the temperature adaptability and efficiency fluctuation are positively correlated with the low-temperature performance coefficient.

3. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method for obtaining the electric ship cluster in step S3 includes: S31: Filtering multiple reference low-temperature performance coefficients based on the change trend and distribution characteristics of the low-temperature performance coefficients of all electric ships; S32: Sequentially obtain ships corresponding to the low-temperature performance coefficients within the range of every two reference low-temperature performance coefficients to form an electric ship cluster.

4. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 3 is characterized in that: The method for screening the reference low-temperature performance coefficient in step S31 includes: S311: Construct a descending sequence of low-temperature performance coefficients for all electric ships; S312: Obtaining a low-temperature performance coefficient difference sequence of the low-temperature performance coefficient descending sequence, and obtaining multiple maximum value points of the low-temperature performance coefficient difference sequence; S313: The low-temperature coefficient of performance corresponding to the maximum point, the maximum low-temperature coefficient of performance, and the minimum low-temperature coefficient of performance are used as reference low-temperature coefficients of performance.

5. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method for obtaining the initial degree of polymerization in step S4 includes: S41: obtaining the average value of the low-temperature performance coefficients of all electric ships in each electric ship cluster as the overall performance level; S42: Sort the electric ship clusters from largest to smallest according to the overall performance levels, and obtain the performance concentration level of each electric ship cluster according to the difference characteristics between the overall performance level of each electric ship cluster and the adjacent overall performance levels and the central trend of the overall performance levels of all electric ship clusters, wherein the difference characteristics are positively correlated with the performance concentration level, and the central trend is negatively correlated with the performance concentration level; S43: Calculate the ratio of the number of electric ships in each electric ship cluster to the number of all electric ships as an aggregation factor; S44: Fusing the performance concentration degree and aggregation factor of each electric ship cluster to obtain the initial aggregation degree of each ship cluster at each sampling moment.

6. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 5 is characterized in that: The method for obtaining the maximum value point in step S42 is: The AMPD algorithm is used to obtain the maximum point of the low-temperature performance coefficient difference sequence.

7. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method for obtaining the updated aggregation degree in step S5 includes: S51: Obtain the mean of the initial aggregation degree of the electric ship cluster at all sampling moments as the overall aggregation level; S52: Obtain an updated aggregation degree of each electric ship cluster according to the overall aggregation level of each electric ship cluster and the change in the initial aggregation degree between adjacent sampling moments.

8. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 7 is characterized in that: Step S52 includes: S521: Selecting the adjacent sampling moments corresponding to the changes that are not equal to the preset changes as reference adjacent sampling moments; S522: Obtain a first ratio between the overall aggregation level and the amount of change in the initial aggregation degree between each reference adjacent sampling moment, and calculate the cumulative sum of the overall aggregation level and the first ratios between all reference adjacent sampling moments as the updated aggregation degree of each ship cluster.

9. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method for obtaining the target aggregated ship cluster in step S6 is: The updated aggregation degree of each ship cluster is normalized. If the normalized result is greater than or equal to the preset aggregation threshold, the corresponding ship cluster is determined to be the target aggregation ship cluster.

10. The dynamic aggregation method of electric ship virtual power station based on low temperature working conditions according to claim 1 is characterized in that: The method further comprises: The charging and discharging strategies of the virtual power station are dynamically adjusted according to the low-temperature performance coefficient of the target aggregated ship cluster, and the ship cluster with a high low-temperature performance coefficient is preferentially dispatched to participate in grid regulation.

Images