Energy storage battery SOC estimation method and system based on charging and discharging current

Abstract

The invention relates to the technical field of energy storage battery electrical performance testing, in particular to an energy storage battery SOC estimation method and system based on charging and discharging current, which is applied to a lithium ion battery, and comprises the following steps: acquiring battery state time sequence data of the energy storage battery in a current sampling time period and a previous sampling time period, and then acquiring abnormal characteristics of each sampling time period; according to the abnormal characteristics of the two sampling time periods, the error factor of the current sampling time period is obtained, and finally the data point weight in the particle filtering algorithm corresponding to the current sampling time period is adjusted by combining the abnormal data characteristics in the two current sampling time periods, so that the accurate adjustment of the weight is realized. The interference of noise data can be reduced, and the precision of particle set acquisition in a particle filtering algorithm is improved, so that the accuracy of SOC estimation of the energy storage battery is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage battery electrical performance testing, and in particular to a method and system for estimating the SOC of an energy storage battery based on charge and discharge current. Background Art

[0002] As a common energy storage battery, lithium-ion battery is usually used to estimate the SOC (State of Charge) of energy storage battery, especially the SOC of energy storage battery with charging and discharging current at the station level. The method used is to estimate the SOC based on particle filter algorithm. However, in the existing SOC estimation process of energy storage battery based on particle filter algorithm, it is too dependent on particle set, and the new particle set is obtained by resampling the weight of particles obtained from the previous particle set through the weight retention and elimination method. Since resampling will make particles with high weights be retained and particles with low weights be eliminated, particles with low weights may be eliminated as noise interference, which will eventually lead to more particles with large weights being retained. Moreover, since the particle set depends on the previous particle set to update, when the data of a certain time period is abnormal or mutated, the accuracy of the new particle set obtained through the previous particle set will decrease due to inaccurate weight confirmation, thereby affecting the accuracy of SOC estimation of energy storage battery. Summary of the invention

[0003] In view of this, in order to solve the technical problem of inaccurate weight acquisition in a particle filter algorithm during SOC estimation based on the particle filter algorithm, the present invention provides a method and system for estimating the SOC of an energy storage battery based on charge and discharge current.

[0004] The technical solutions adopted are as follows: In a first aspect, an embodiment of the present invention provides a method for estimating the SOC of an energy storage battery based on charge and discharge current, comprising: Acquire battery status time series data of the energy storage battery in the current sampling time period and the previous sampling time period, wherein the battery status time series data at least includes current; the current is a charging current or a discharging current; Based on the battery status time series data, an abnormal feature sequence of each sampling time period is obtained, wherein the abnormal feature sequence includes abnormal features of each sampling moment in the sampling time period, and the abnormal feature is an abnormal data feature or a normal data feature; According to the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period, and the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period, an error factor of the current sampling time period is obtained, and the error factor is used to characterize the degree of weight deviation of the data points in the current sampling time period; According to the difference in the isolation state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

[0005] In combination with the first aspect above, in some possible implementations, the battery status time series data further includes battery temperature and battery internal resistance; Based on the battery status time series data, an abnormal feature sequence of each sampling time period is obtained, including: According to the current, obtaining the abnormal current state in each sampling time period; According to the abnormal current state, the battery temperature and the battery internal resistance, an abnormal feature sequence in each sampling time period is obtained.

[0006] In combination with the first aspect above, in some possible implementations, the process of acquiring the abnormal current state in the sampling time period includes: Respectively obtaining the current difference between each sampling moment and its two adjacent sampling moments within the sampling time period to obtain a first current difference and a second current difference; Obtaining an absolute value of a difference between the first current difference and the second current difference to obtain a first current abnormality degree at each sampling moment; Obtaining an average value of the absolute value of the first current difference within a sampling period to obtain a second current abnormality degree; According to the first current abnormality degree and the second current abnormality degree, the current abnormality state at each sampling moment in the sampling time period is obtained.

[0007] In combination with the first aspect, in some possible implementations, obtaining an abnormal feature sequence in each sampling time period according to the abnormal current state, the battery temperature, and the battery internal resistance includes: According to the battery temperature, obtaining the battery temperature fluctuation degree and the battery temperature extreme difference in a sampling time period; According to the battery temperature and the battery internal resistance, obtaining a correlation between changes in the battery temperature and the battery internal resistance during a sampling period; According to the abnormal current state, the battery temperature fluctuation degree, the battery temperature extreme difference and the change correlation, an abnormal feature sequence of the sampling time period is obtained; wherein the abnormal current state, the battery temperature fluctuation degree and the battery temperature extreme difference are positively correlated with the abnormal features of the abnormal feature sequence, and the change correlation is negatively correlated with the abnormal features.

[0008] In combination with the first aspect above, in some possible implementations, the energy storage battery SOC estimation method further includes: Clustering the abnormal features in the abnormal feature sequence to obtain a plurality of clusters; Based on the overall numerical level of abnormal features in each cluster, the plurality of clusters are divided into abnormal clusters and normal clusters, the abnormal features in the abnormal clusters are the abnormal data features, and the abnormal features in the normal clusters are the normal data features.

[0009] In combination with the first aspect above, in some possible implementations, the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period is specifically a difference in quantity proportion, and the quantity proportion difference is a difference between the quantity proportion of the abnormal data features in the abnormal feature sequence of the current sampling time period and the quantity proportion of the abnormal data features in the abnormal feature sequence of the previous sampling time period; The correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period is specifically the overall level of the centroid distance, which is the average value of the distance between the centroid of each normal cluster in the current sampling time period and the centroid of the abnormal cluster in the current sampling time period; The error factor of the current sampling time period is obtained by the difference in the number of proportions and the overall level of the centroid distance, wherein the error factor and the difference in the number of proportions are positively correlated with the overall level of the centroid distance.

[0010] In combination with the first aspect, in some possible implementations, according to the difference in the isolation state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted, including: Obtaining the isolated state of the abnormal cluster of the current sampling time period in the abnormal feature sequence of the current sampling time period to obtain a first isolated state, and obtaining the isolated state of the abnormal cluster of the previous sampling time period in the abnormal feature sequence of the previous sampling time period to obtain a second isolated state; comparing the first isolated state and the second isolated state; According to the comparison result between the first isolated state and the second isolated state and in combination with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

[0011] In combination with the first aspect, in some possible implementations, the process of acquiring the first isolated state includes: Obtain the distance between the centroid of the abnormal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as the first abnormal distance, and obtain the distance between the centroid of each normal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as each first normal distance, calculate the sum of the first abnormal distance and all the first normal distances, and obtain a first total distance; taking the ratio of the first abnormal distance to the first total distance as the first isolated state; The process of acquiring the second isolated state includes: Obtain the distance between the centroid of the abnormal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as the second abnormal distance, and obtain the distance between the centroid of each normal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as each second normal distance, calculate the sum of the second abnormal distance and all second normal distances, and obtain the second total distance; The ratio of the second abnormal distance to the second total distance is taken as the second isolated state.

[0012] In combination with the first aspect, in some possible implementations, adjusting the weight of the data point in the particle filter algorithm corresponding to the current sampling time period according to the comparison result between the first isolated state and the second isolated state and in combination with the error factor includes: If the first isolated state is smaller than the second isolated state, the adjusted data point weights of the sampling moments corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: The adjusted data point weights of the sampling time corresponding to the normal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: If the first isolated state is greater than the second isolated state, the adjusted data point weights of the sampling moments corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: The adjusted data point weights of the sampling time corresponding to the normal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: in, is the initial data point weight corresponding to the rth abnormal data feature in the current sampling time period, is the adjusted data point weight corresponding to the rth abnormal data feature in the current sampling time period, is the initial data point weight corresponding to the sth normal data feature in the current sampling time period, is the adjusted data point weight corresponding to the sth normal data feature in the current sampling time period, and G is the normalized value of the error factor in the current sampling time period.

[0013] In a second aspect, an embodiment of the present invention provides a system for estimating the SOC of an energy storage battery based on charge and discharge current, comprising a memory and a processor, wherein the memory is used to store executable program code, and the processor is used to call and run the program code from the memory to implement the above-mentioned method for estimating the SOC of an energy storage battery based on charge and discharge current.

[0014] The present invention has the following technical effects, including but not limited to: according to the battery state time series data of the two adjacent sampling time periods, namely the current sampling time period and the previous sampling time period of the energy storage battery, the abnormal feature sequence of each sampling time period is obtained, so as to obtain the specific classification of the abnormal features at each sampling moment in the sampling time period, whether it belongs to abnormal data features or normal data features, thereby improving the accuracy of abnormal data analysis; then, according to the difference of abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period, and the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period, the error factor of the current sampling time period is obtained, and the error factor is used to characterize the degree of deviation of the weight of the data point in the current sampling time period. The error factor is a very important parameter in the subsequent weight adjustment process. By obtaining an accurate error factor that reflects the characteristics of the current sampling time period, the accuracy of the subsequent weight adjustment can be improved; finally, according to the difference of the isolated state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted. Therefore, the method provided by the present invention can accurately adjust the weights of the data points in the particle filter algorithm corresponding to the current sampling time period according to the abnormal state of the data in the current sampling time period and the association with the previous sampling time period, which can reduce the interference of noise data and improve the accuracy of particle set acquisition in the particle filter algorithm, thereby improving the accuracy of SOC estimation of the energy storage battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a hardware composition diagram of a storage battery SOC estimation device based on charge and discharge current; Figure 2 It is a flow chart of a method for estimating SOC of an energy storage battery based on charge and discharge current provided by the present invention; Figure 3 is a flow chart of the implementation process of step S2; Figure 4 is a flowchart of the implementation process of step S2-1; Figure 5 is a flowchart of the implementation process of step S2-2; Figure 6 It is a flow chart of the process of dividing abnormal data features and normal data features; Figure 7 is a flow chart of the implementation process of step S4; Figure 8 It is a structural schematic diagram of a storage battery SOC estimation system based on charge and discharge current; Fig. 9 It is a schematic diagram of the structure of a computer-readable storage medium; Among them, 1 is a battery status data acquisition module, and 2 is a data processor. DETAILED DESCRIPTION

[0016] The device used in the energy storage battery SOC estimation method based on charge and discharge current (hereinafter referred to as the energy storage battery SOC estimation method) provided in this embodiment is an energy storage battery SOC estimation device based on charge and discharge current. The application object of the energy storage battery SOC estimation device - the energy storage battery is specifically a lithium-ion battery, and is used to estimate the SOC of the lithium-ion battery.

[0017] like Figure 1 As shown, the energy storage battery SOC estimation device includes a battery status data acquisition module 1 and a data processor 2. The battery status data acquisition module 1 and the data processor 2 are connected by signals, which can be connected by wired signal lines or by wireless communication modules.

[0018] The battery status data acquisition module 1 is used to obtain the battery status time series data of the energy storage battery. The battery status data acquisition module 1 can be used as an external detection device for the energy storage battery, or it can be integrated with the energy storage battery. It should be understood that a sampling time period is set, and the battery status data acquisition module 1 is used to obtain the battery status time series data of the energy storage battery within the sampling time period. The length of the sampling time period is set according to actual needs, such as: 10 seconds. Multiple sampling moments are set within the sampling time period, and the time interval between adjacent sampling moments (i.e., the sampling period) is set according to actual needs, such as 0.1 seconds. The battery status data acquisition module 1 obtains the battery status data of the energy storage battery at each sampling moment, and the battery status time series data includes the battery status data at each sampling moment. Taking the above example, the sampling time period includes the battery status data at 100 sampling moments.

[0019] The data processor 2 is used to process the acquired battery status time series data to adjust the weight of the data points. The hardware device corresponding to the data processor 2 can be a computer host or server device with certain data processing capabilities.

[0020] Data processor 2 is used to execute Figure 2 The energy storage battery SOC estimation method shown in FIG. Figure 2 As shown, the energy storage battery SOC estimation method includes the following steps: Step S1: obtaining battery status time series data of a current sampling time period and a previous sampling time period of an energy storage battery, wherein the battery status time series data at least includes current; the current is a charging current or a discharging current.

[0021] This embodiment adjusts the weight of the data points in the adjacent next sampling time period in the particle filter algorithm according to the data of the previous sampling time period. Therefore, this embodiment sets two sampling time periods, namely the current time period and the previous sampling time period adjacent thereto. It should be understood that the end time of the current time period is the current time.

[0022] The battery state time series data of the current sampling time period and the previous sampling time period of the energy storage battery are obtained by the battery state data acquisition module 1. Since the estimation of the SOC of the energy storage battery is mainly carried out by the change of the current, it is necessary to collect the current data of the energy storage battery. Then, the battery state time series data includes at least the current. Because the estimation of the SOC of the energy storage battery by the particle filter algorithm also needs to rely on other state information of the energy storage battery at each sampling time, in addition to the current, other state data of the energy storage battery can also be collected according to specific needs, such as the temperature of the energy storage battery, the internal resistance of the energy storage battery, the power of the energy storage battery, the voltage of the energy storage battery, etc. Other state data and current data are collected at the same sampling time and sampling time period. In addition, since the SOC of the energy storage battery can be estimated during the charging or discharging process, the current of the energy storage battery is the charging current or the discharging current. The charging current is used for estimation during the charging process, and the discharging current is used for estimation during the discharging process. Correspondingly, the battery state data acquisition module 1 includes a current sensor for detecting the current data of the energy storage battery.

[0023] Step S2: Based on the battery status time series data, an abnormal feature sequence of each sampling time period is obtained, wherein the abnormal feature sequence includes abnormal features of each sampling moment in the sampling time period, and the abnormal features are abnormal data features or normal data features.

[0024] Since the battery status time series data characterizes the battery status at each sampling moment of the energy storage battery, the abnormal feature sequence of each sampling time period of the energy storage battery can be obtained according to the battery status time series data of each sampling moment. For any sampling time period in the current sampling time period and the previous sampling time period, the abnormal feature sequence includes the abnormal features of each sampling moment in the sampling time period, and the abnormal features characterize the abnormal battery conditions of the energy storage battery at the corresponding sampling moment. Since the abnormal conditions include abnormal conditions and normal conditions, the abnormal features are abnormal data features or normal data features. The abnormal data features characterize that the energy storage battery is in a relatively abnormal state at the corresponding sampling moment, and the normal data features characterize that the energy storage battery is in a relatively normal state at the corresponding sampling moment.

[0025] Step S3: According to the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period, and the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period, the error factor of the current sampling time period is obtained, and the error factor is used to characterize the degree of deviation of the data point weights in the current sampling time period.

[0026] When the energy storage battery is abnormal, there will be a certain difference between the abnormal feature sequence of the current sampling time period and the previous sampling time period, that is, the difference between the abnormal feature sequence of the current sampling time period and the previous sampling time period will characterize the abnormal state of the energy storage battery in the current sampling time period, and the obtained abnormal state will affect the adjustment of the weight of the data point corresponding to the current sampling time period. Moreover, the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period also characterizes the abnormal state of the energy storage battery in the current sampling time period in one aspect. Therefore, according to the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period, and the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period, the error factor of the current sampling time period can be obtained. The error factor is used to characterize the degree of deviation of the weight of the data point in the current sampling time period, and can also be understood as the possibility of the existence of the weight deviation of the data point in the current sampling time period. The error factor will ultimately determine the degree of adjustment of the weight of the data point.

[0027] Step S4: According to the difference in the isolation state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

[0028] The isolated state of the abnormal data feature in the corresponding abnormal feature sequence in the current sampling time period and any sampling time period in the previous sampling time period reflects the abnormal situation of the energy storage battery state in the corresponding sampling time period to a certain extent. According to the difference in the isolated state of the abnormal data feature in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor obtained in step S3, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period can be adjusted.

[0029] In a specific embodiment, the battery status time series data in step S1 includes the battery temperature and the battery internal resistance in addition to the current of the energy storage battery, that is, the battery status data at each sampling moment includes the battery temperature and the battery internal resistance in addition to the current. The battery temperature can be detected by a temperature sensor arranged on the energy storage battery, and the battery internal resistance can be detected by a detection circuit for detecting the battery internal resistance. In order to realize SOC estimation, the battery status data can also include the power information, voltage information, etc. of the energy storage battery.

[0030] Energy storage batteries, especially field-level energy storage batteries, are required to have faster charging and discharging speeds, so the corresponding energy storage batteries have higher power density and energy density, that is, field-level energy storage batteries have very high charging and discharging efficiency. Higher charging and discharging currents will speed up the process, but may cause the battery's internal resistance to increase and heat to increase, which in turn affects efficiency. Low currents may cause the charging and discharging speeds to slow down, and the more stable the current, the higher its efficiency. The higher the stability, the less the data is affected by noise. Because changes in battery temperature will affect the rate of chemical reactions inside the battery, the more drastic the current changes, the more drastic the current fluctuations, and the worse the accuracy of the collected current data. Therefore, based on the changes in current, combined with the relationship between the battery temperature and the changes in the battery's internal resistance, the abnormal characteristics of each sampling moment are obtained. In a specific embodiment, such as Figure 3 As shown, the implementation process of step S2 is as follows: Step S2-1: According to the current, obtain the abnormal current state in each sampling time period.

[0031] Step S2-2: Obtain an abnormal feature sequence for each sampling time period according to the abnormal current state, the battery temperature and the battery internal resistance.

[0032] In a specific embodiment, Figure 4 As shown, the implementation process of step S2-1 is as follows: Step S2-1-1: respectively obtain the current difference between each sampling moment and its two adjacent sampling moments within the sampling time period to obtain a first current difference and a second current difference.

[0033] For the convenience of explanation, any sampling time period between the current sampling time period and the previous sampling time period is referred to as a target sampling time period, and any sampling moment within the target sampling time period is referred to as a target sampling moment.

[0034] Under normal circumstances, the current data at each sampling moment is similar to the current data at the previous and next adjacent sampling moments. The greater the difference between the current data at each sampling moment and the current data at the previous and next adjacent sampling moments, the more abnormal the current data at this moment is. The analysis here is aimed at the difference in current changes caused by the change of current state in different charging and discharging periods. This situation will be considered to be caused by noise, and will eventually be given a lower weight and eliminated during the resampling process. In fact, the state change result is not caused by noise, so it should be given a larger weight and retained.

[0035] The current difference between the target sampling moment and the adjacent previous sampling moment in the target sampling time period is set as the first current difference, and the current difference between the target sampling moment and the adjacent next sampling moment in the target sampling time period is set as the second current difference. In this embodiment, the first current difference is the difference between the current at the target sampling moment and the current at the adjacent previous sampling moment, and the second current difference is the difference between the current at the adjacent next sampling moment and the current at the target sampling moment. Therefore, the current difference is the difference between the current at the next sampling moment and the current at the previous sampling moment in two adjacent sampling moments, so the current difference will have positive and negative points. Therefore, the current difference can reflect the change trend and degree of change of the two adjacent sampling moments. If the current difference is positive, and the larger the value, it means that the current at the next sampling moment is greater than the current at the previous sampling moment, and the gap between the two is larger; if the current difference is negative, and the larger the value, it means that the current at the next sampling moment is less than the current at the previous sampling moment, and the gap between the two is larger.

[0036] Step S2-1-2: Obtain the absolute value of the difference between the first current difference and the second current difference to obtain the degree of first current abnormality at each sampling moment.

[0037] The absolute value of the difference between the first current difference and the second current difference is used as the first current abnormality degree at the target sampling moment, thereby obtaining the first current abnormality degree at each sampling moment. The greater the first current abnormality degree, the greater the possibility that the current at the target sampling moment is abnormal.

[0038] Step S2-1-3: Obtain an average value of the absolute value of the first current difference within a sampling time period to obtain a second current abnormality degree.

[0039] Since the target sampling time period includes multiple sampling moments, it also includes multiple first current differences. If the target sampling time period is set to include N sampling moments, it includes N-1 first current differences. Obtain the absolute value of each first current difference in the target sampling time period to remove the positive and negative. Then calculate the average value of the absolute value of the first current difference in the target sampling time period, and the result is the second current abnormality degree. The second current abnormality degree is based on the overall characteristics of the target sampling time period. The reason for obtaining the second current abnormality degree is that the faster the current changes in the target sampling time period, the greater the possibility of abnormality in its state. Therefore, the stability and abnormality of the current are further quantified by the change difference of the current data at all adjacent moments. The greater the second current abnormality degree, the worse the stability of the current in the target sampling time period, and the greater the possibility of abnormality; conversely, the smaller the possibility of abnormality. The characteristics of the current at each sampling moment are reflected by characterizing the overall current characteristics of the target sampling time period.

[0040] Step S2-1-4: According to the first current abnormality degree and the second current abnormality degree, the current abnormality state at each sampling moment in the sampling time period is obtained.

[0041] The calculation formula of the current abnormal state at the target sampling time in the target sampling time period is as follows: in, represents the abnormal current state at the i-th sampling moment in the target sampling time period, represents the first current abnormality degree at the i-th sampling moment in the target sampling time period, Indicates the second current abnormality level during the target sampling period.

[0042] By adopting the above process, the abnormal current state at each sampling moment in the current sampling time period and the previous sampling time period is obtained.

[0043] In a specific embodiment, Figure 5 As shown, the implementation process of step S2-2 is specifically as follows: Step S2-2-1: According to the battery temperature, obtain the battery temperature fluctuation degree and the battery temperature extreme difference in the sampling time period.

[0044] The increase in the temperature of the energy storage battery will reduce the internal resistance of the battery and increase the current. However, if the temperature is too high and reaches a certain high temperature quickly, it may cause instability and lead to current fluctuations. Therefore, the abnormal fluctuation of the current at this time is caused by its own fluctuation rather than noise interference. Therefore, it is necessary to analyze and correct the abnormality of the status information at each sampling moment through the specific trend of the change of internal resistance under temperature.

[0045] The target sampling time period includes the battery temperature at each sampling time, and the battery temperature fluctuation degree of the target sampling time period is obtained. The battery temperature fluctuation degree can be characterized by the temperature variance. Therefore, in this embodiment, the variance of the battery temperature in the target sampling time period is obtained, and the obtained variance is used as the battery temperature fluctuation degree of the target sampling time period. The maximum battery temperature and the minimum battery temperature in the target sampling time period are obtained, and the difference between the maximum battery temperature and the minimum battery temperature is calculated to obtain the battery temperature extreme difference of the target sampling time period.

[0046] The greater the battery temperature fluctuation, the more drastic the temperature fluctuation; the greater the battery temperature range in the target sampling time period, the more drastic the temperature change, and the more abnormal the battery status data in the target sampling time period.

[0047] Step S2-2-2: According to the battery temperature and the battery internal resistance, obtain the change correlation between the battery temperature and the battery internal resistance in the sampling time period.

[0048] Since the battery temperature will affect the internal resistance only when it reaches a certain level, and because the change of the battery internal resistance is related to the temperature and the characteristics of the battery material, in addition to being related to the temperature. Therefore, when the temperature reaches a certain high temperature, the sudden change of the internal resistance will reduce the accuracy of the energy storage battery estimation, that is, the stronger the correlation between the temperature change and the internal resistance change, the higher the accuracy of the energy storage battery estimation through the temperature state information data, and the weaker the change correlation, the greater the error in the energy storage battery estimation process. Accordingly, according to the battery temperature and battery internal resistance of the target sampling time period, the change correlation of the battery temperature and battery internal resistance of the target sampling time period is obtained. As a specific implementation, the battery temperature at each moment of the target sampling time period is normalized, and the battery temperature sequence is obtained according to the normalized battery temperature at each moment. Similarly, the battery internal resistance at each moment of the target sampling time period is normalized, and the battery internal resistance sequence is obtained according to the normalized battery internal resistance at each moment. Obtain the correlation between the battery temperature sequence and the battery internal resistance sequence. As a specific implementation, obtain the Pearson correlation coefficient between the battery temperature sequence and the battery internal resistance sequence. The smaller the Pearson correlation coefficient is, the greater the error will be in the energy storage battery SOC estimation process, and the more likely it is to have an abnormality. As another implementation, the correlation can also be calculated by cosine similarity.

[0049] Step S2-2-3: According to the abnormal current state, the battery temperature fluctuation degree, the battery temperature extreme difference and the change correlation, obtain the abnormal characteristic sequence of the sampling time period; wherein, the abnormal current state, the battery temperature fluctuation degree and the battery temperature extreme difference are positively correlated with the abnormal characteristics of the abnormal characteristic sequence, and the change correlation is negatively correlated with the abnormal characteristics.

[0050] According to the abnormal current state, battery temperature fluctuation degree, battery temperature extreme difference and change correlation obtained in the above steps S-2-1 and S-2-2, the abnormal feature sequence of the target sampling time period is obtained. The abnormal feature sequence is composed of the abnormal features of each sampling moment in the target sampling time period. Moreover, it can be seen from the above steps S-2-1 and S-2-2 that the abnormal current state, battery temperature fluctuation degree and battery temperature extreme difference are positively correlated with the abnormal features of the abnormal feature sequence, and the change correlation is negatively correlated with the abnormal features.

[0051] As a specific implementation, a specific quantification method of abnormal characteristics is given as follows: in, represents the abnormal characteristics of the i-th sampling time in the target sampling time period, It represents the normalized result of battery temperature fluctuation. It represents the normalized result of the battery temperature extreme difference. represents the correlation of changes, exp represents the exponential function with the natural constant e as the base, It should be understood that the normalization method in this embodiment can be a maximum and minimum value normalization method, or can be the following normalization method: , where y is the normalized value and x is the input value.

[0052] By adopting the above method, the abnormal characteristics of each sampling moment in the current sampling time period and the previous sampling time period are obtained. The abnormal characteristics indicate the abnormal degree of the battery status data at the corresponding sampling moment. The larger the abnormal characteristics, the more abnormal it is, and the smaller the abnormal characteristics, the more normal it is. Thus, the abnormal characteristic sequence of the current sampling time period and the abnormal characteristic sequence of the previous sampling time period are obtained.

[0053] In a specific embodiment, the energy storage battery SOC estimation method further includes a process of dividing abnormal data features and normal data features, such as Figure 6 As shown, the details are as follows: Step S2-3: Cluster the abnormal features in the abnormal feature sequence to obtain several clusters. Among them, the abnormal features at each sampling moment in the target sampling time period are taken as the vertical axis, and the sampling moment is taken as the horizontal axis to construct a two-dimensional coordinate system. Based on the two-dimensional coordinate system, the abnormal feature space of the target sampling time period can be constructed, and the abnormal features of all sampling moments in the target sampling time period are mapped in the abnormal feature space. This embodiment can adopt an existing clustering method, such as a density-based DBSCAN clustering method, or a K-Means clustering algorithm for clustering. If the K-Means clustering algorithm is adopted, the number of clusters can be set manually or obtained by the elbow method. The DBSCAN clustering method, the K-Means clustering algorithm and the elbow method are all well-known technologies and will not be described in detail. As other implementation methods, several thresholds of different sizes can also be set, and the abnormal features are divided into multiple intervals according to the thresholds to achieve division into multiple clusters. It should be understood that no matter what clustering method is adopted, the abnormal features in the abnormal feature sequence need to be divided into at least two clusters.

[0054] Thus, several clusters of abnormal feature sequences in the current sampling time period and several clusters of abnormal feature sequences in the previous sampling time period are obtained.

[0055] Step S2-4: Based on the overall numerical level of the abnormal features in each cluster, the plurality of clusters are divided into abnormal clusters and normal clusters, the abnormal features in the abnormal clusters are the abnormal data features, and the abnormal features in the normal clusters are the normal data features.

[0056] Since there is a monotonic relationship between the degree of abnormality represented by the abnormal feature and the value of the abnormal feature, the larger the value, the higher the degree of abnormality. Therefore, the overall numerical level of the abnormal features in each cluster is obtained, wherein the overall numerical level can be represented by the average value of the abnormal features in the cluster. According to the overall numerical level of the abnormal features of each cluster, several clusters are divided into abnormal clusters and normal clusters. As a specific implementation, the cluster corresponding to the largest overall numerical level is determined as an abnormal cluster, and each cluster other than the abnormal cluster is determined as a normal cluster. Then, the abnormal features in the abnormal cluster are abnormal data features, and the abnormal features in the normal cluster are normal data features.

[0057] By adopting the above method, the abnormal clusters and normal clusters of the abnormal feature sequence in the current sampling time period and the abnormal clusters and normal clusters of the abnormal feature sequence in the previous sampling time period are obtained.

[0058] Since the battery power data at all sampling moments in the current sampling time period is predicted and analyzed through the battery power data at all sampling moments in the previous sampling time period, the predicted observed power of the particle corresponding to each sampling moment is obtained through the difference between the predicted power data and the actual observed power, and the weight of each particle is analyzed through the difference between the actual observed power and the predicted observed power of each particle, and a new particle set is obtained by resampling according to the weight of the particle. Therefore, the greater the difference in the abnormal characteristic distribution of the battery status data in the current sampling time period and the previous sampling time period, the lower the accuracy of the particle set in the current sampling time period obtained through the particle set in the previous sampling time period, resulting in a decrease in the accuracy of the subsequent energy storage battery SOC estimation.

[0059] In step S3, the greater the distribution difference between the abnormal clusters in the current sampling time period and the previous sampling time period, the more it indicates that there is a process of change in the state of the energy storage battery in these two sampling time periods. Therefore, when the battery state data in the current sampling time period is resampled by the battery state data in the previous sampling time period, the battery state data with a large change in the current sampling time period will be discarded, resulting in a certain error.

[0060] As a specific implementation, in step S3, the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period is specifically a quantity proportion difference, and the process of obtaining the quantity proportion difference is as follows: obtaining the quantity proportion of the abnormal data features in the abnormal feature sequence of the current sampling time period, and obtaining the quantity proportion of the abnormal data features in the abnormal feature sequence of the previous sampling time period, and then calculating the difference between the two quantity proportions to obtain the quantity proportion difference. A specific calculation formula is provided as follows: in, Indicates the difference in quantity proportion, Indicates the total number of sampling moments in the current sampling time period, which is the same as the total number of sampling moments in the previous sampling time period; Indicates the number of abnormal data features in the abnormal cluster in the previous sampling period, Indicates the number of abnormal data features in the abnormal cluster within the current sampling period; It indicates the difference in the number proportion. The larger the difference in the number proportion, the greater the difference in particle state information between the current sampling period and the previous sampling period. That is, the greater the error in estimating the next particle set by the previous particle set, the greater the possibility of error.

[0061] In step S3, the association between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period is specifically the overall level of the centroid distance. The process of obtaining the overall level of the centroid distance is as follows: obtain the distance between the centroid of each normal cluster of the current sampling time period and the centroid of the abnormal cluster of the current sampling time period, and then calculate the average value of the obtained distance, which is the overall level of the centroid distance. Therefore, the overall level of the centroid distance characterizes the association between each normal cluster of the current sampling time period and the abnormal cluster of the current sampling time period. The higher the overall level of the centroid distance, the less associated the normal clusters of the current sampling time period are with the abnormal clusters of the current sampling time period, and the larger the value of the overall level of the centroid distance, the greater the possibility of error.

[0062] The error factor of the current sampling time period in step S3 is obtained by the above-obtained quantity proportion difference and the overall level of the centroid distance, wherein the error factor is positively correlated with the quantity proportion difference and the overall level of the centroid distance. A specific calculation formula is provided as follows: in, Indicates the error factor of the current sampling period, that is, the error factor between the current sampling period and the previous sampling period. Indicates the overall level of the centroid distance corresponding to the current sampling time period.

[0063] In a specific embodiment, Figure 7 As shown, the implementation process of step S4 is as follows; Step S4-1: Obtain the isolation state of the abnormal cluster of the current sampling time period in the abnormal feature sequence of the current sampling time period to obtain a first isolation state, and obtain the isolation state of the abnormal cluster of the previous sampling time period in the abnormal feature sequence of the previous sampling time period to obtain a second isolation state.

[0064] Among them, the isolated state can be explained as: the deviation between the position of the abnormal cluster in the corresponding abnormal feature sequence and the central position of the corresponding abnormal feature sequence. The larger the deviation, the more the abnormal cluster deviates from the central position of the abnormal feature sequence and the more isolated it is.

[0065] In this embodiment, the process of obtaining the first isolated state includes: obtaining the distance between the centroid of the abnormal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as the first abnormal distance, and obtaining the distance between the centroid of each normal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as each first normal distance, calculating the sum of the first abnormal distance and all first normal distances, and obtaining the first total distance. The ratio of the first abnormal distance to the first total distance is taken as the first isolated state. It should be understood that the centroid of the abnormal cluster, the centroid of the normal cluster, the center point of the abnormal feature sequence (i.e., the center point of the abnormal feature space), and the distance between the centroid and the center point are all obtained by mapping to a two-dimensional coordinate system.

[0066] Similarly, the process of obtaining the second isolated state includes: obtaining the distance between the centroid of the abnormal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as the second abnormal distance, and obtaining the distance between the centroid of each normal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as each second normal distance, calculating the sum of the second abnormal distance and all second normal distances to obtain the second total distance. The ratio of the second abnormal distance to the second total distance is used as the second isolated state.

[0067] Step S4-2: Compare the first isolated state and the second isolated state.

[0068] Step S4-3: According to the comparison result between the first isolated state and the second isolated state, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

[0069] Specifically, if the first isolated state is smaller than the second isolated state, it means that the abnormal cluster is closer to the center point of all data points in the abnormal feature sequence in the current sampling time period than in the previous sampling time period (indicating that the data in the abnormal cluster in the current sampling time period is likely to be abnormal data caused by a change in state. At this time, the data in the abnormal cluster is not caused by noise, so it needs to be retained). At this time, the data points in the abnormal cluster may be retained as noise through the particle filter algorithm. Therefore, the distance between the abnormal data points and the center point in the current sampling time period should be reduced, that is, the weight of the data points in the abnormal cluster should be increased to retain these abnormal points; at the same time, the weight of the data points in the normal cluster should be reduced so that the data in these normal clusters are retained. Then, the adjusted data point weights of the sampling time corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: in, is the initial data point weight corresponding to the rth abnormal data feature in the current sampling time period (that is, obtained through the particle filter algorithm), is the adjusted data point weight corresponding to the rth abnormal data feature in the current sampling time period, and G is the normalized value of the error factor in the current sampling time period, which is calculated specifically above. .

[0070] The adjusted data point weights at the sampling time corresponding to the normal data features in the particle filter algorithm in the current sampling time period are as follows: in, is the initial data point weight corresponding to the sth normal data feature in the current sampling period, is the adjusted data point weight corresponding to the sth normal data feature in the current sampling time period.

[0071] On the contrary, if the first isolated state is greater than the second isolated state, it means that the abnormal cluster is farther away from the center point of all data points in the abnormal feature sequence in the current sampling time period than in the previous sampling time period (indicating that the data in the abnormal cluster in the current sampling time period is less likely to be abnormal data caused by a change in state. At this time, the data in the abnormal cluster is likely to be caused by noise, so it does not need to be retained). The distance between the abnormal data points and the center point in the current sampling time period should be increased, that is, the weight of the data points in the abnormal cluster should be reduced to discard these abnormal points. Then, the adjusted data point weights of the sampling time corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: The adjusted data point weights at the sampling time corresponding to the normal data features in the particle filter algorithm in the current sampling time period are as follows: It should be understood that if the first isolated state is equal to the second isolated state, the weight is not adjusted.

[0072] Through the above process, the corrected weight of each sampling moment in the current sampling time period, that is, each data point, that is, each particle is obtained.

[0073] It should be understood that two adjacent sampling time periods are not limited in time to the current sampling time period and the previous sampling time period. The focus of this application is to adjust the weight of the subsequent sampling time period according to the previous sampling time period. As time changes, the "current sampling time period" will also become the "previous sampling time period" of its next sampling time period.

[0074] Based on the corrected weight of each particle, the example filtering algorithm is used to estimate the SOC of the energy storage battery. This part belongs to the prior art, such as a power battery SOC estimation method based on an extended Kalman particle filter algorithm disclosed in the authorization announcement number CN103472398B, or a battery state of charge estimation method disclosed in the publication number CN104573401A. In this embodiment, the corrected observation value and the predicted value of the particle are obtained by weight analysis, and the energy storage battery SOC estimation is completed by the corrected weight. Specifically: the corrected weight of each particle is normalized (that is, the sum is 1); the particle set is resampled (that is, sampling with replacement) by the normalized weight to obtain a new particle set. Through the new particle set until the particle set corresponding to the current sampling time period is traversed, the energy storage battery SOC estimation is completed by the weight of the current corresponding particle set and the battery status information.

[0075] This embodiment also provides a storage battery SOC estimation system 700 based on charge and discharge current, such as Figure 8 As shown, it includes a memory 701 and a processor 702, the memory 701 is used to store an executable program code 703, and the processor 702 is used to call and run the program code 703 from the memory 701 to implement Figure 2 The energy storage battery SOC estimation method based on charging and discharging current is shown.

[0076] This embodiment also provides a computer-readable storage medium 800, such as Fig. 9 As shown, the computer readable storage medium 800 stores a computer program code 801. When the computer program code 801 is executed on a computer, the computer executes Figure 2 The energy storage battery SOC estimation method based on charging and discharging current is shown.

[0077] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for estimating the SOC of an energy storage battery based on charge and discharge current, characterized in that: include: Acquire battery status time series data of the energy storage battery in the current sampling time period and the previous sampling time period, wherein the battery status time series data at least includes current; the current is a charging current or a discharging current; Based on the battery status time series data, an abnormal feature sequence of each sampling time period is obtained, wherein the abnormal feature sequence includes abnormal features of each sampling moment in the sampling time period, and the abnormal feature is an abnormal data feature or a normal data feature; According to the difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period, and the correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period, an error factor of the current sampling time period is obtained, and the error factor is used to characterize the degree of weight deviation of the data points in the current sampling time period; According to the difference in the isolation state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

2. The energy storage battery SOC estimation method based on charge and discharge current according to claim 1, characterized in that: The battery status time series data also includes battery temperature and battery internal resistance; Based on the battery status time series data, an abnormal feature sequence of each sampling time period is obtained, including: According to the current, obtaining the abnormal current state in each sampling time period; According to the abnormal current state, the battery temperature and the battery internal resistance, an abnormal feature sequence in each sampling time period is obtained.

3. The energy storage battery SOC estimation method based on charge and discharge current according to claim 2, characterized in that: The process of obtaining the abnormal current state in the sampling time period includes: Respectively obtaining the current difference between each sampling moment and its two adjacent sampling moments within the sampling time period to obtain a first current difference and a second current difference; Obtaining an absolute value of a difference between the first current difference and the second current difference to obtain a first current abnormality degree at each sampling moment; Obtaining an average value of the absolute value of the first current difference within a sampling period to obtain a second current abnormality degree; According to the first current abnormality degree and the second current abnormality degree, the current abnormality state at each sampling moment in the sampling time period is obtained.

4. The energy storage battery SOC estimation method based on charge and discharge current according to claim 2, characterized in that: According to the abnormal current state, the battery temperature and the battery internal resistance, an abnormal feature sequence of each sampling time period is obtained, including: According to the battery temperature, obtaining the battery temperature fluctuation degree and the battery temperature extreme difference in a sampling time period; According to the battery temperature and the battery internal resistance, obtaining a correlation between changes in the battery temperature and the battery internal resistance during a sampling period; According to the abnormal current state, the battery temperature fluctuation degree, the battery temperature extreme difference and the change correlation, an abnormal feature sequence of the sampling time period is obtained; wherein the abnormal current state, the battery temperature fluctuation degree and the battery temperature extreme difference are positively correlated with the abnormal features of the abnormal feature sequence, and the change correlation is negatively correlated with the abnormal features.

5. The energy storage battery SOC estimation method based on charge and discharge current according to claim 1, characterized in that: The energy storage battery SOC estimation method further includes: Clustering the abnormal features in the abnormal feature sequence to obtain a plurality of clusters; Based on the overall numerical level of abnormal features in each cluster, the plurality of clusters are divided into abnormal clusters and normal clusters, the abnormal features in the abnormal clusters are the abnormal data features, and the abnormal features in the normal clusters are the normal data features.

6. The method for estimating SOC of a storage battery based on charge and discharge current according to claim 5, characterized in that: The difference between the abnormal data features in the abnormal feature sequence of the current sampling time period and the previous sampling time period is specifically a difference in quantity proportion, and the quantity proportion difference is the difference between the quantity proportion of the abnormal data features in the abnormal feature sequence of the current sampling time period and the quantity proportion of the abnormal data features in the abnormal feature sequence of the previous sampling time period; The correlation between the abnormal data features and the normal data features in the abnormal feature sequence of the current sampling time period is specifically the overall level of the centroid distance, which is the average value of the distance between the centroid of each normal cluster in the current sampling time period and the centroid of the abnormal cluster in the current sampling time period; The error factor of the current sampling time period is obtained by the difference in the number of proportions and the overall level of the centroid distance, wherein the error factor and the difference in the number of proportions are positively correlated with the overall level of the centroid distance.

7. The energy storage battery SOC estimation method based on charge and discharge current according to claim 5 or 6, characterized in that: According to the difference in the isolation state of the abnormal data features in the current sampling time period and the previous sampling time period in the corresponding abnormal feature sequence, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted, including: Obtaining the isolated state of the abnormal cluster of the current sampling time period in the abnormal feature sequence of the current sampling time period to obtain a first isolated state, and obtaining the isolated state of the abnormal cluster of the previous sampling time period in the abnormal feature sequence of the previous sampling time period to obtain a second isolated state; comparing the first isolated state and the second isolated state; According to the comparison result between the first isolated state and the second isolated state and in combination with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted.

8. The method for estimating the SOC of an energy storage battery based on charge and discharge current according to claim 7, characterized in that: The process of acquiring the first isolated state includes: Obtain the distance between the centroid of the abnormal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as the first abnormal distance, and obtain the distance between the centroid of each normal cluster in the current sampling time period and the center point of the abnormal feature sequence in the current sampling time period as each first normal distance, calculate the sum of the first abnormal distance and all the first normal distances, and obtain a first total distance; taking the ratio of the first abnormal distance to the first total distance as the first isolated state; The process of acquiring the second isolated state includes: Obtain the distance between the centroid of the abnormal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as the second abnormal distance, and obtain the distance between the centroid of each normal cluster in the previous sampling time period and the center point of the abnormal feature sequence in the previous sampling time period as each second normal distance, calculate the sum of the second abnormal distance and all second normal distances, and obtain the second total distance; The ratio of the second abnormal distance to the second total distance is taken as the second isolated state.

9. The method for estimating the SOC of an energy storage battery based on charge and discharge current according to claim 8, characterized in that: According to the comparison result between the first isolated state and the second isolated state, combined with the error factor, the weight of the data point in the particle filter algorithm corresponding to the current sampling time period is adjusted, including: If the first isolated state is smaller than the second isolated state, the adjusted data point weights of the sampling moments corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: The adjusted data point weights of the sampling time corresponding to the normal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: If the first isolated state is greater than the second isolated state, the adjusted data point weights of the sampling moments corresponding to the abnormal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: The adjusted data point weights of the sampling time corresponding to the normal data features in the particle filter algorithm corresponding to the current sampling time period are as follows: in, is the initial data point weight corresponding to the rth abnormal data feature in the current sampling time period, is the adjusted data point weight corresponding to the rth abnormal data feature in the current sampling time period, is the initial data point weight corresponding to the sth normal data feature in the current sampling time period, is the adjusted data point weight corresponding to the sth normal data feature in the current sampling time period, and G is the normalized value of the error factor in the current sampling time period.

10. A system for estimating the SOC of an energy storage battery based on charge and discharge current, characterized in that: It includes a memory and a processor, wherein the memory is used to store executable program code, and the processor is used to call and run the program code from the memory to implement the energy storage battery SOC estimation method based on charge and discharge current as described in any one of claims 1-9.

Images