Method for monitoring energy storage state of energy storage power station

Abstract

The invention relates to an energy storage state monitoring method for an energy storage power station, and belongs to the technical field of energy storage. The method comprises the steps of collecting basic data of an energy storage power station; preprocessing the basic data of the power station; determining sample data and labels corresponding to the sample data according to the initial data, and then performing division to construct a training set and a test set; using the training set to train the LSTM model, and using the test set to test to obtain a temperature prediction model; and predicting in real time, and generating countermeasure suggestions according to a prediction result. In the process of training the temperature prediction model, the position of the battery in the battery pack and the influence of the position on the battery temperature are considered, and the confidence coefficient of the training data is determined, so that the performance of the temperature prediction model is improved, and the accuracy of the prediction result output by the temperature prediction model can be improved. Therefore, the accuracy of monitoring the energy storage state of the energy storage power station can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy storage, and in particular relates to a method for monitoring the energy storage status of an energy storage power station. Background Art

[0002] At present, energy storage power stations have problems such as battery safety risks, battery consistency management, insufficient system protection, aging, and maintenance challenges. The details are as follows:

[0003] Battery safety hazards: Batteries and other energy storage devices used in battery energy storage power stations have safety hazards. Lithium-ion batteries use low-boiling-point, flammable organic electrolytes, and the material system has a high calorific value. After the battery body or electrical equipment fails, it is easy to trigger the exothermic side reaction of the battery material, causing thermal runaway of the battery, which may evolve into major safety accidents such as combustion and explosion of the energy storage system.

[0004] Battery consistency management issues: There are many batteries in energy storage power stations, which are arranged relatively densely, and the number of single cells can reach tens of thousands. Battery consistency management becomes a problem. If a single cell fails due to abuse and thermal runaway, it is very easy to cause a chain reaction in the surrounding batteries, increasing safety risks.

[0005] Insufficient system protection: The safety protection system of the energy storage power station may be imperfect. For example, if the information sharing between the battery management system (BMS), the power storage converter (PCS) and the energy management system (EMS) is incomplete or untimely, the fault may not be effectively controlled in a timely manner and may evolve into an accident.

[0006] Aging problem: As the battery is used for a longer time, its performance will gradually decline, including reduced capacity and increased internal resistance. These problems may cause the battery to generate more heat during charging and discharging, increasing the risk of thermal runaway.

[0007] Maintenance challenges: Due to the large number of batteries and their dense arrangement in energy storage power stations, regular maintenance and inspection of batteries is a huge challenge. If maintenance is not timely or in place, it may increase the risk of battery failure and accidents.

[0008] In addition, the current state monitoring of energy storage power stations has problems such as low accuracy, low reliability, and low stability. Therefore, how to overcome the shortcomings of existing technologies is an urgent problem to be solved in the field of energy storage technology. Summary of the invention

[0009] The purpose of the present invention is to solve the deficiencies of the prior art and to provide a method for monitoring the energy storage status of an energy storage power station.

[0010] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0011] The method for monitoring the energy storage status of an energy storage power station comprises the following steps:

[0012] S1. Collect basic data of energy storage power stations, including battery performance parameters, battery operation parameters, and environmental data;

[0013] The energy storage power station includes multiple battery packs, each of which includes multiple energy storage batteries; battery performance parameters include: capacity, voltage, resistance and heat generation rate; battery operating parameters include: cycle life, discharge depth and operating temperature;

[0014] S2. Preprocess the basic data of the energy storage power station to obtain the initial data of the battery;

[0015] The preprocessing is to normalize the battery performance parameters and battery operation parameters collected by S1, and then construct the initial data of the battery [A, B, C, D, E, F, G, H, T];

[0016] Among them, A represents the normalized value of the battery capacity; B represents the normalized value of the variance of the battery voltage sequence at the sampling time; C represents the normalized value of the battery resistance; E represents the normalized value of the battery cycle life; F represents the normalized value of the depth of discharge; G represents the distance between the battery and the adjacent battery; H represents the number of batteries in the battery pack where the battery is located; T represents the operating temperature of the battery at the sampling time;

[0017] S3. Determine sample data and its corresponding labels based on the initial data, and then divide them to construct a training set and a test set; wherein each sample data is [A, B, C, D, G, H] and the label is [E, F, T];

[0018] S4. The LSTM model is trained using the training set, with sample data as input and corresponding labels as output, and the test set is used for testing to obtain a temperature prediction model;

[0019] S5. Preprocess the basic data of the energy storage power station currently collected in real time and input it into the temperature prediction model constructed in S4 to perform prediction, thereby obtaining prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature.

[0020] Further, preferably, in S1, the environmental data includes the spacing distance between every two batteries, the number of batteries included in each battery pack, and the temperature of the batteries during operation.

[0021] Furthermore, preferably, in S3, the sample ratio of the training set and the test set is 8:2.

[0022] Furthermore, preferably, in S4, during training, a back propagation algorithm is used to update the model parameters.

[0023] Furthermore, preferably, in S4, during training, the loss function used is as follows:

[0024]

[0025] Among them, E i represents the normalized value of the cycle life of the ith battery, F i Represents the normalized value of the discharge depth of the ith battery, T i represents the temperature of the ith battery at the sampling time; X i represents the normalized value of the predicted cycle life of the ith battery, Y i Represents the normalized value of the predicted discharge depth of the ith battery, Z i represents the predicted temperature of the ith battery; R i represents the first confidence of the i-th battery; S i Represents the second confidence of the i-th battery.

[0026] Further, preferably, the first confidence of the battery is calculated by: determining the distance between the battery and the center of the battery pack, recorded as the first distance; determining the diagonal distance of the battery pack; calculating the ratio of the first distance and the diagonal distance, and the ratio is the first confidence of the battery.

[0027] Further, preferably, the second confidence is calculated according to the capacity difference, the resistance difference and the voltage difference; the average of the capacity difference, the resistance difference and the voltage difference of the battery is calculated, and the reciprocal of the average is determined as the second confidence of the battery;

[0028] The difference between the capacity of each battery and other batteries in the same battery pack is calculated; the mean of all the difference means is calculated; the difference between the difference mean and the mean of all the difference means is calculated; the quotient of the difference and the mean of all the difference means is calculated to obtain the capacity difference of each battery in the battery pack;

[0029]

[0030] in, is the capacity difference corresponding to the i-th battery, A ij is the difference between the capacity of the ith battery and the jth battery, is the absolute value of the difference between the capacity of the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0031] Calculate the mean difference value of the resistance of each battery in the same battery pack and other batteries; calculate the mean of all the mean difference values; calculate the difference between the mean difference value and the mean of all the mean difference values; calculate the quotient of the difference value and the mean of all the mean difference values ​​to obtain the resistance difference of each battery in the battery pack;

[0032]

[0033] in, is the resistance difference corresponding to the ith battery, B ij is the difference in resistance between the ith battery and the jth battery, is the absolute value of the difference in resistance between the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0034] Calculate the mean of the difference between the voltage of each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values ​​to obtain the voltage difference of each battery in the battery pack;

[0035]

[0036] in, is the voltage difference corresponding to the ith battery, C ij is the difference between the voltages of the ith battery and the jth battery, is the absolute value of the voltage difference between the ith battery and the jth battery; n is the number of batteries in the battery pack.

[0037] Furthermore, preferably, the method also includes converting the prediction results into specific quantitative indicators of hidden dangers, specifically: pre-dividing the level of cycle life, the level of discharge depth, and the level of predicted temperature, and then taking the highest level among the level of predicted cycle life, the level of predicted discharge depth, and the level of predicted temperature as the hidden danger level.

[0038] Furthermore, preferably, it further comprises:

[0039] S6. Generate countermeasures and suggestions based on the prediction results; the countermeasures and suggestions include monitoring suggestions and maintenance suggestions.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] High temperature will accelerate the electrochemical reaction rate inside the battery, resulting in an increase in the rate at which electrical energy is converted into heat energy, which in turn exacerbates the temperature rise inside the battery and causes the stability of the output current of the energy storage power station to decrease. In order to improve the maintenance efficiency of the energy storage power station, it is necessary to monitor the working status of the power station in a timely manner according to the temperature inside the power station, and perform maintenance in a timely manner when problems such as abnormal temperature occur.

[0042] At present, the temperature of the energy storage power station is usually monitored in real time based on the temperature sensor in the energy storage power station. However, this state monitoring method has the problems of poor real-time performance and low efficiency. When the temperature received by the sensor is abnormal, the battery in the energy storage power station may have been damaged, resulting in low maintenance efficiency of the energy storage power station.

[0043] In order to solve this problem, the prediction model can be trained based on the attribute data of different battery groups and different batteries in the energy storage power station, as well as the voltage and current data during operation, to predict the temperature of different batteries in the energy storage power station. In this way, the future temperature changes can be predicted based on the current operating data of the battery at work, which can improve the timeliness of temperature monitoring, and thus improve the efficiency of status monitoring and maintenance. In the process of training the model, the confidence of the training data corresponding to the battery is evaluated based on the data such as the capacitance and voltage of the battery at work and the distance between the battery and the center of the battery group, thereby improving the performance of the model, which can improve the accuracy of the predicted temperature, and thus improve the accuracy of monitoring the status of the energy storage power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a schematic diagram of neurons in the temperature prediction model;

[0045] Figure 2 It is a structural diagram of the temperature prediction model. DETAILED DESCRIPTION

[0046] The present invention is further described in detail below in conjunction with embodiments.

[0047] Those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific techniques or conditions are specified in the examples, the techniques or conditions described in the literature in the art or the product specifications are used. If the manufacturer of the materials or equipment used is not specified, they are all conventional products that can be purchased.

[0048] Example 1

[0049] The method for monitoring the energy storage status of an energy storage power station comprises the following steps:

[0050] S1. Collect basic data of energy storage power stations, including battery performance parameters, battery operation parameters, and environmental data;

[0051] The energy storage power station includes multiple battery packs, each of which includes multiple energy storage batteries; battery performance parameters include: capacity, voltage, resistance and heat generation rate; battery operating parameters include: cycle life, discharge depth and operating temperature;

[0052] S2. Preprocess the basic data of the energy storage power station to obtain the initial data of the battery;

[0053] The preprocessing is to normalize the battery performance parameters and battery operation parameters collected by S1, and then construct the initial data of the battery [A, B, C, D, E, F, G, H, T];

[0054] Among them, A represents the normalized value of the battery capacity; B represents the normalized value of the variance of the battery voltage sequence at the sampling time; C represents the normalized value of the battery resistance; E represents the normalized value of the battery cycle life; F represents the normalized value of the depth of discharge; G represents the distance between the battery and the adjacent battery; H represents the number of batteries in the battery pack where the battery is located; T represents the operating temperature of the battery at the sampling time;

[0055] S3. Determine sample data and its corresponding labels based on the initial data, and then divide them to construct a training set and a test set; wherein each sample data is [A, B, C, D, G, H] and the label is [E, F, T];

[0056] S4. The LSTM model is trained using the training set, with sample data as input and corresponding labels as output, and the test set is used for testing to obtain a temperature prediction model;

[0057] S5. Preprocess the basic data of the energy storage power station currently collected in real time and input it into the temperature prediction model constructed in S4 to perform prediction, thereby obtaining prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature.

[0058] Example 2

[0059] The method for monitoring the energy storage status of an energy storage power station comprises the following steps:

[0060] S1. Collect basic data of energy storage power stations, including battery performance parameters, battery operation parameters, and environmental data;

[0061] The energy storage power station includes multiple battery packs, each of which includes multiple energy storage batteries; battery performance parameters include: capacity, voltage, resistance and heat generation rate; battery operating parameters include: cycle life, discharge depth and operating temperature;

[0062] S2. Preprocess the basic data of the energy storage power station to obtain the initial data of the battery;

[0063] The preprocessing is to normalize the battery performance parameters and battery operation parameters collected by S1, and then construct the initial data of the battery [A, B, C, D, E, F, G, H, T];

[0064] Among them, A represents the normalized value of the battery capacity; B represents the normalized value of the variance of the battery voltage sequence at the sampling time; C represents the normalized value of the battery resistance; E represents the normalized value of the battery cycle life; F represents the normalized value of the depth of discharge; G represents the distance between the battery and the adjacent battery; H represents the number of batteries in the battery pack where the battery is located; T represents the operating temperature of the battery at the sampling time;

[0065] S3. Determine sample data and its corresponding labels based on the initial data, and then divide them to construct a training set and a test set; wherein each sample data is [A, B, C, D, G, H] and the label is [E, F, T];

[0066] S4. The LSTM model is trained using the training set, with sample data as input and corresponding labels as output, and the test set is used for testing to obtain a temperature prediction model;

[0067] S5. Preprocess the basic data of the energy storage power station currently collected in real time and input it into the temperature prediction model constructed in S4 to perform prediction, thereby obtaining prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature.

[0068] In S1, the environmental data includes the spacing distance between every two batteries, the number of batteries contained in each battery pack, and the temperature of the batteries during operation.

[0069] In S3, the sample ratio of the training set and the test set is 8:2.

[0070] In S4, during training, the back propagation algorithm is used to update the model parameters.

[0071] In S4, the loss function used during training is as follows:

[0072]

[0073] Among them, E i represents the normalized value of the cycle life of the ith battery, F i Represents the normalized value of the discharge depth of the ith battery, T i represents the temperature of the ith battery at the sampling time; X i represents the normalized value of the predicted cycle life of the ith battery, Y i Represents the normalized value of the predicted discharge depth of the ith battery, Z i represents the predicted temperature of the ith battery; R i represents the first confidence of the i-th battery; S i Represents the second confidence of the i-th battery.

[0074] The first confidence level of the battery is calculated as follows: determine the distance between the battery and the center of the battery pack, recorded as the first distance; determine the diagonal distance of the battery pack; calculate the ratio of the first distance and the diagonal distance, and the ratio is the first confidence level of the battery.

[0075] Calculating a second confidence level according to the capacity difference, the resistance difference, and the voltage difference; calculating the average of the capacity difference, the resistance difference, and the voltage difference of the battery, and determining the reciprocal of the average as the second confidence level of the battery;

[0076] The difference between the capacity of each battery and other batteries in the same battery pack is calculated; the mean of all the difference means is calculated; the difference between the difference mean and the mean of all the difference means is calculated; the quotient of the difference and the mean of all the difference means is calculated to obtain the capacity difference of each battery in the battery pack;

[0077]

[0078] in, is the capacity difference corresponding to the i-th battery, A ij is the difference between the capacity of the ith battery and the jth battery, is the absolute value of the difference between the capacity of the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0079] Calculate the mean difference value of the resistance of each battery in the same battery pack and other batteries; calculate the mean of all the mean difference values; calculate the difference between the mean difference value and the mean of all the mean difference values; calculate the quotient of the difference value and the mean of all the mean difference values ​​to obtain the resistance difference of each battery in the battery pack;

[0080]

[0081] in, is the resistance difference corresponding to the ith battery, B ij is the difference in resistance between the ith battery and the jth battery, is the absolute value of the difference in resistance between the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0082] Calculate the mean of the difference between the voltage of each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values ​​to obtain the voltage difference of each battery in the battery pack;

[0083]

[0084] in, is the voltage difference corresponding to the ith battery, Cij is the difference between the voltages of the ith battery and the jth battery, is the absolute value of the voltage difference between the ith battery and the jth battery; n is the number of batteries in the battery pack.

[0085] The method also includes converting the prediction results into specific quantitative indicators of hidden dangers, specifically: pre-dividing the level of cycle life, the level of discharge depth, and the level of predicted temperature, and then taking the highest level among the level of predicted cycle life, the level of predicted discharge depth, and the level of predicted temperature as the hidden danger level.

[0086] Also includes:

[0087] S6. Generate countermeasures and suggestions based on the prediction results; the countermeasures and suggestions include monitoring suggestions and maintenance suggestions.

[0088] Example 3

[0089] The method for monitoring the energy storage status of an energy storage power station comprises the following steps:

[0090] S1, collects basic data of energy storage power stations, including battery performance parameters, battery operation parameters, and environmental data;

[0091] In one embodiment of the present application, the energy storage power station includes multiple battery groups, each battery group includes multiple energy storage batteries, and the basic data of the power station includes: battery performance parameters, battery operating parameters and environmental data.

[0092] The performance parameters of the battery include at least: capacity, voltage, resistance, and heat generation rate.

[0093] The operating parameters of the battery include at least: cycle life, discharge depth, and operating temperature.

[0094] The environmental data of the battery includes at least: the distance between every two batteries, the number of batteries contained in each battery pack, and the temperature of the battery during operation. The above data are inherent to the energy storage battery and can be obtained through the battery technical manual or actual testing.

[0095] Specifically, among battery performance parameters, the larger the capacity, the more electricity the battery can store and the longer it can provide. Since batteries with lower capacity are charged and discharged more frequently, and batteries may reach their life limit faster and fail under frequent charging and discharging, the lower the capacity, the lower the stability of the battery's performance.

[0096] The stability of voltage can be used to characterize the health of the battery. The greater the voltage fluctuation, the higher the probability of battery failure. The smaller the voltage fluctuation, the lower the probability of battery failure. Specifically, the voltage data of the energy storage battery can be collected according to a preset sampling frequency, for example, once every 0.1 seconds, to obtain a voltage data sequence; the variance of the voltage sequence data is calculated, and the variance is used to characterize the voltage volatility. The larger the variance, the greater the volatility, and the smaller the variance, the smaller the volatility.

[0097] The smaller the battery resistance, the smaller the energy loss of the battery during the charging and discharging process, the higher the efficiency, and the lower the probability of failure; the higher the battery resistance, the higher the degree of aging or damage of the battery, and therefore the higher the probability of battery failure.

[0098] Different types of batteries (such as lithium-ion batteries, lead-acid batteries, etc.) generate different amounts of heat during the charging and discharging process. The heat generation rate of each battery can be confirmed so that the temperature change of the battery can be accurately predicted. The heat generation rate is the amount of heat generated by the energy storage battery per unit time, and its unit can be watts (W) or joules per second (J / s). Specifically, the higher the heat generation rate of the energy storage battery, the higher the heat generated by the energy storage battery per unit time, and the higher the probability of failure.

[0099] Specifically, among the operating parameters of the battery, the cycle life of the battery is a fixed number of charge and discharge cycles, which is used to characterize the number of times the battery can be cycled and discharged, and can be used to evaluate the durability of the battery. When the number of charge and discharge cycles of the battery is closer to its cycle life, the probability of battery failure is higher. Therefore, the ratio of the current number of charge and discharge cycles of the battery to the cycle life can be calculated. The higher the ratio, the higher the probability of battery failure.

[0100] The depth of discharge (DOD) of a battery is used to characterize the ratio of the battery's used capacity to its total capacity. The higher the ratio, the more the battery's calculation formula satisfies the following relationship:

[0101] DOD = (1-current remaining power / total battery power) × 100%

[0102] In the above formula, the total battery capacity refers to the capacity of the battery when it is fully charged, and the current remaining capacity refers to the remaining capacity of the battery. For example, when the capacity of any battery is 100Ah (ampere-hours) and the current remaining capacity is 20Ah, its depth of discharge is: DOD = (1-20Ah / 100Ah) × 100% = 80%. The higher the depth of discharge, the more chemical reactions will occur inside the battery, thereby accelerating the aging and capacity decay of the battery. The battery life may be shorter and the probability of battery failure is higher.

[0103] Environmental data: Record the environmental data of the energy storage power station, such as the distance between each two batteries, the number of batteries in each battery pack, and the temperature of the battery during operation. These conditions have an important impact on the performance and safety of the battery. Among them, the higher the distance between each two batteries, the more timely the heat dissipation of the battery, and the lower the probability of battery failure; the higher the number of batteries in each battery pack, the higher the density of battery distribution, and the higher the probability of battery failure; the higher the temperature of the battery during operation, the higher the probability of battery failure.

[0104] S2, preprocessing the basic data of the power station to obtain initial data.

[0105] In an embodiment of the present application, different data may have dimension differences, so each type of data may be normalized to eliminate the dimension differences. For example, a linear normalization algorithm or a mean algorithm may be used, which is not limited in the present application.

[0106] In one embodiment of the present application, for any battery, the initial data of the battery may be in the form of a vector, for example: [A, B, C, D, E, F, G, H, T]. Wherein A represents the normalized value of the battery capacity, B represents the normalized value of the variance of the battery voltage sequence at the sampling time, C represents the normalized value of the battery resistance, E represents the normalized value of the battery cycle life, F represents the normalized value of the depth of discharge, G represents the distance between the battery and the adjacent battery, H represents the number of batteries in the battery pack where the battery is located, and T represents the temperature of the battery at the sampling time.

[0107] S3, determine the sample data and its corresponding label based on the initial data, and then divide it to construct a training set and a test set; in one embodiment of the present application, for each battery, the sample and label can be determined based on the initial data corresponding to the battery, and a sample and its corresponding label are determined as a piece of training data. Specifically, the label data is the cycle life, discharge depth, and temperature of the battery at the sampling time, and the sample data is other initial data except temperature. For example, the sample data is [A, B, C, D, G, H], and the label is [E, F, T].

[0108] In an embodiment of the present application, 80% of the training data may be used as a training set, and 20% of the training data may be used as a training set.

[0109] S4, train the LSTM model using the training set, take the sample data as input, take the corresponding label as output, and use the test set for testing to obtain the temperature prediction model;

[0110] In one embodiment of the present application, sample data in the training set can be input into a pre-built initial prediction model, and the output is a label corresponding to each battery for training; wherein the data in the label includes a normalized value of the battery's cycle life, a normalized value of the depth of discharge, and the temperature of the battery at the sampling time;

[0111] The loss value of the initial prediction model is calculated based on the actual label data and the predicted label data; when the loss value is greater than the preset termination threshold, the parameters in the initial prediction model are updated based on the back propagation algorithm until the loss value is less than or equal to the termination threshold, thereby obtaining a temperature prediction model trained to a convergent state.

[0112] In one embodiment of the present application, when calculating the loss value of the initial prediction model based on the label data and the predicted label, the attention mechanism can be used to adjust the loss value, thereby improving the accuracy of the calculated loss value.

[0113] Specifically, since energy storage batteries are thermally sensitive, the sensitivity of batteries to temperature needs to be considered when analyzing the thermal management scheme of energy storage batteries. Thermal sensitivity means that the increase in temperature will accelerate the aging of energy storage batteries, reduce the performance of batteries, and even cause safety accidents; while too low a temperature will affect the charging and discharging performance of batteries. That is, the higher the temperature of the energy storage battery, the higher the probability of an accident. Since the heat accumulation problem of batteries closer to the center of the battery pack is more serious, the closer to the center of the battery pack, the lower the first confidence of the battery; the farther away from the center of the battery pack, the higher the first confidence of the battery. Among them, the first confidence is used to characterize the credibility of the probability of battery failure. The calculation method of the first confidence of the battery is: determine the distance between the battery and the center of the battery pack, recorded as the first distance; determine the diagonal distance of the battery pack; calculate the ratio of the first distance to the diagonal distance, and the ratio is the first confidence of the battery.

[0114] In addition, the confidence of each energy storage battery can be determined based on the consistency of the energy storage batteries in the energy storage battery pack. The consistency of energy storage batteries refers to the degree to which each battery in a battery pack composed of multiple batteries maintains relative consistency in key parameters such as capacity, resistance, and voltage. The consistency of energy storage batteries in the same battery pack has a relatively important impact on the overall performance, service life, and safety of the battery pack. Specifically, the consistency of energy storage batteries mainly includes the following aspects:

[0115] Battery capacity consistency: refers to the degree of capacity difference of each battery in the same battery pack. When the capacity difference of energy storage batteries in the same energy storage battery pack does not exceed a preset ratio (such as 5%), it indicates that the battery pack can work evenly during the charging and discharging process to avoid overcharging or over-discharging of certain batteries, thereby affecting the performance and life of the entire battery pack. Specifically, calculate the mean of the difference values ​​between the capacity of each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values, and obtain the capacity difference of each battery in the battery pack;

[0116]

[0117] in, is the capacity difference corresponding to the i-th battery, A ij is the difference between the capacity of the ith battery and the jth battery, is the absolute value of the difference between the capacity of the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0118] Battery resistance consistency: Resistance is a measure of the internal resistance of a battery, and is used to characterize the energy loss of a battery during the charge and discharge process. The degree of resistance difference between individual batteries in the same battery pack is lower than a preset ratio (such as 10%), indicating that the heat loss energy of each single battery in the battery pack during discharge is relatively consistent, thereby maintaining the overall energy state of the battery pack. Specifically, calculate the mean of the difference values ​​of the resistance between each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values, and obtain the resistance difference of each battery in the battery pack.

[0119]

[0120] in, is the resistance difference corresponding to the ith battery, B ij is the difference in resistance between the ith battery and the jth battery, is the absolute value of the difference in resistance between the ith battery and the jth battery; n is the number of batteries in the battery pack;

[0121] Battery voltage consistency: Voltage is one of the important parameters of battery performance, which can affect the output power and stability of the battery. When the voltage difference between each battery in the same battery pack does not exceed a preset ratio (such as 5%), it can avoid voltage imbalance in parallel or series use, which may lead to degradation or damage to the battery pack performance. Specifically, calculate the mean of the voltage difference between each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values ​​to obtain the voltage difference of each battery in the battery pack.

[0122]

[0123] in, is the voltage difference corresponding to the ith battery, C ij is the difference between the voltages of the ith battery and the jth battery, is the absolute value of the voltage difference between the ith battery and the jth battery; n is the number of batteries in the battery pack.

[0124] In one embodiment of the present application, the second confidence level can be calculated based on the capacity difference, resistance difference, and voltage difference; the higher the second confidence level, the higher the credibility of the data of the energy storage battery. Specifically, the mean of the capacity difference, resistance difference, and voltage difference of any battery can be calculated, and the reciprocal of the mean can be determined as the second confidence level of the battery. For example, when the mean is 0.2, the second confidence level is 5.

[0125]

[0126] The beneficial effects of ensuring the consistency of energy storage batteries include: Improving battery pack performance: Battery packs with good consistency can work more evenly and reduce the overall performance degradation caused by differences in the performance of single cells. Extending service life: Battery packs with good consistency can prevent some batteries from failing prematurely due to overcharging or over-discharging, thereby extending the service life of the entire battery pack. Improving safety: Battery packs with good consistency can reduce safety hazards such as thermal runaway and short circuits caused by differences in the performance of single cells.

[0127] In an embodiment of the present application, calculating the loss value of the initial prediction model according to the label data and the predicted label includes:

[0128]

[0129] Among them, E i represents the normalized value of the cycle life of the ith battery, F i Represents the normalized value of the discharge depth of the ith battery, T i represents the temperature of the ith battery at the sampling time; Xi represents the normalized value of the predicted cycle life of the ith battery, Y i Represents the normalized value of the predicted discharge depth of the ith battery, Z i represents the predicted temperature of the ith battery; R i represents the first confidence of the i-th battery; S i Represents the second confidence of the i-th battery.

[0130] S5, pre-processing the basic data of the energy storage power station currently collected in real time, and inputting it into the temperature prediction model constructed in S4 to perform prediction, thereby obtaining prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature;

[0131] The prediction model is preferably a long short-term memory model, which includes neurons that are repeatedly used in series, and the structure of the neurons is as follows: Figure 1 As shown, the model structure is as Figure 2 As shown,

[0132] Figure 1 In the equation, Xt is the sample data input at time t;

[0133] ht represents the prediction result, i.e. the label, which is obtained by combining Ct and ot. Please refer to the figure below for the formula;

[0134] ft is the forgotten data, and its value range is 0 to 1. The way to obtain ft includes: calculating Xt and ht-1 (the prediction result at time t-1) based on the sigmoid function to obtain a value between 0 and 1. The closer the forgotten data ft is to 0, the more predicted labels predicted by the first sample data at the previous moment need to be forgotten. The closer the forgotten data is to 1, the fewer predicted labels predicted by the first sample data at the previous moment need to be forgotten. Specifically, Wf and bf are the weight and bias corresponding to the forgotten data ft.

[0135] it is the input data. The methods of obtaining it include: calculating Xt and ht-1 (the prediction result at time t-1) based on the sigmoid function to obtain a value between 0 and 1; Wi and bi are the weights and biases corresponding to the input data.

[0136] is the initial state value of the neuron at time t; Wc and bc are the weights and biases corresponding to the initial state value.

[0137] ot is the hidden output data at time t. The method of obtaining ot includes: calculating Xt and ht-1 (the prediction result at time t-1) based on the sigmoid function to obtain a value between 0 and 1; Wo and bo are the weights and biases corresponding to the hidden output data;

[0138] Ct represents the state value of the neuron at time t.

[0139] exist Figure 2 In the figure, although multiple neurons are shown, there is actually only one neuron, which will be used repeatedly. For example, when training a temperature prediction model, the basic data at time t and the basic data at time t+1 can be collected, the values ​​of the initial state and the initial prediction results can be randomly set in advance, and the neuron corresponding to time t can be input to obtain the prediction result at time t and the output state data; at time t+1, the input of the neuron becomes the prediction result at time t, the state data output at time t, and the basic data at time t+1, and this cycle is repeated until the training is completed. The loss value is determined based on the prediction result at time t and the cycle life, discharge depth, and temperature in the basic data at time t; the loss value at time t+1 is determined based on the prediction result at time t+1 and the cycle life, discharge depth, and temperature in the basic data at time t+1.

[0140] In one embodiment of the present application, basic data of energy storage batteries collected in real time are preprocessed based on a temperature prediction model and then input into the temperature prediction model to obtain prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature; and risk assessment of batteries in energy storage power stations is performed based on the prediction results.

[0141] Specifically, the prediction results can be converted into specific hidden danger quantitative indicators, such as hidden danger level, hidden danger quantity, hidden danger distribution, etc. The specific conversion method is as follows:

[0142] Exemplarily, the levels of cycle life, discharge depth, and predicted temperature are pre-classified. Specifically, the cycle life can be determined as level 1 for 1-3 months, level 2 for 3-6 months, etc.; the discharge depth can be determined as level 1 for 0-20%, level 2 for 20%-30%, etc.; the predicted temperature can be determined as level 1 for 25 degrees Celsius-28 degrees Celsius, and level 2 for 28 degrees Celsius-30 degrees Celsius. The level of predicted cycle life is determined according to the classification method of cycle life; the level of predicted discharge depth is determined according to the classification method of discharge depth; and the level of predicted temperature is determined according to the classification method of temperature.

[0143] The highest level among the predicted cycle life level, the predicted discharge depth level and the predicted temperature level can be used as the hidden danger level; the average value of the predicted cycle life level, the predicted discharge depth level and the predicted temperature level can also be used as the hidden danger level.

[0144] S6, generating countermeasures and suggestions based on the prediction results.

[0145] Strengthen monitoring: Improve the monitoring level of energy storage power stations, monitor the operating status and environmental conditions of batteries in real time, and take immediate measures to deal with any abnormalities.

[0146] Optimized maintenance: Develop a reasonable maintenance plan based on the prediction results and potential danger analysis. Regularly inspect, test and maintain the battery to ensure battery performance and safety.

[0147] In some embodiments of the present application, the prediction results can also be used to generate a prediction report: based on the prediction results, a battery safety hazard prediction report of the energy storage power station is compiled. The report should include prediction methods, data sources, prediction results, hazard analysis, countermeasures and suggestions, etc.

[0148] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for monitoring the energy storage status of an energy storage power station, characterized in that: The steps include: S1. Collect basic data of energy storage power stations, including battery performance parameters, battery operation parameters, and environmental data; The energy storage power station includes multiple battery packs, each of which includes multiple energy storage batteries; battery performance parameters include: capacity, voltage, resistance and heat generation rate; battery operating parameters include: cycle life, discharge depth and operating temperature; S2. Preprocess the basic data of the energy storage power station to obtain the initial data of the battery; The preprocessing is to normalize the battery performance parameters and battery operation parameters collected by S1, and then construct the initial data of the battery [A, B, C, D, E, F, G, H, T]; Among them, A represents the normalized value of the battery capacity; B represents the normalized value of the variance of the battery voltage sequence at the sampling time; C represents the normalized value of the battery resistance; E represents the normalized value of the battery cycle life; F represents the normalized value of the depth of discharge; G represents the distance between the battery and the adjacent battery; H represents the number of batteries in the battery pack where the battery is located; T represents the operating temperature of the battery at the sampling time; S3. Determine sample data and its corresponding labels based on the initial data, and then divide them to construct a training set and a test set; wherein each sample data is [A, B, C, D, G, H] and the label is [E, F, T]; S4. The LSTM model is trained using the training set, with sample data as input and corresponding labels as output, and the test set is used for testing to obtain a temperature prediction model; S5. Preprocess the basic data of the energy storage power station currently collected in real time and input it into the temperature prediction model constructed in S4 to perform prediction, thereby obtaining prediction results; the prediction results include predicted cycle life, predicted discharge depth and predicted temperature.

2. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: In S1, the environmental data includes the spacing distance between every two batteries, the number of batteries contained in each battery pack, and the temperature of the batteries during operation.

3. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: In S3, the sample ratio of the training set and the test set is 8:

2.

4. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: In S4, during training, the back propagation algorithm is used to update the model parameters.

5. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: In S4, the loss function used during training is as follows: Among them, E i represents the normalized value of the cycle life of the ith battery, F i Represents the normalized value of the discharge depth of the ith battery, T i represents the temperature of the ith battery at the sampling time; X i represents the normalized value of the predicted cycle life of the ith battery, Y i Represents the normalized value of the predicted discharge depth of the ith battery, Z i represents the predicted temperature of the ith battery; R i represents the first confidence of the i-th battery; S i Represents the second confidence of the i-th battery.

6. The method for monitoring the energy storage status of an energy storage power station according to claim 5, characterized in that: The first confidence level of the battery is calculated as follows: determine the distance between the battery and the center of the battery pack, recorded as the first distance; determine the diagonal distance of the battery pack; calculate the ratio of the first distance and the diagonal distance, and the ratio is the first confidence level of the battery.

7. The method for monitoring the energy storage status of an energy storage power station according to claim 5, characterized in that: Calculating a second confidence level according to the capacity difference, the resistance difference, and the voltage difference; calculating the average of the capacity difference, the resistance difference, and the voltage difference of the battery, and determining the reciprocal of the average as the second confidence level of the battery; The difference between the capacity of each battery and other batteries in the same battery pack is calculated; the mean of all the difference means is calculated; the difference between the difference mean and the mean of all the difference means is calculated; the quotient of the difference and the mean of all the difference means is calculated to obtain the capacity difference of each battery in the battery pack; in, is the capacity difference corresponding to the i-th battery, A ij is the difference between the capacity of the ith battery and the jth battery, is the absolute value of the difference between the capacity of the ith battery and the jth battery, and n is the number of batteries in the battery pack; Calculate the mean difference value of the resistance of each battery in the same battery pack and other batteries; calculate the mean of all the mean difference values; calculate the difference between the mean difference value and the mean of all the mean difference values; calculate the quotient of the difference value and the mean of all the mean difference values ​​to obtain the resistance difference of each battery in the battery pack; in, is the resistance difference corresponding to the ith battery, B ij is the difference in resistance between the ith battery and the jth battery, is the absolute value of the difference in resistance between the ith battery and the jth battery; n is the number of batteries in the battery pack; Calculate the mean of the difference between the voltage of each battery and other batteries in the same battery pack; calculate the mean of all the mean of the difference values; calculate the difference between the mean of the difference value and the mean of all the mean of the difference values; calculate the quotient of the difference and the mean of all the mean of the difference values ​​to obtain the voltage difference of each battery in the battery pack; in, is the voltage difference corresponding to the ith battery, C ij is the difference between the voltages of the ith battery and the jth battery, is the absolute value of the voltage difference between the ith battery and the jth battery; n is the number of batteries in the battery pack.

8. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: The method also includes converting the prediction results into specific quantitative indicators of hidden dangers, specifically: pre-dividing the level of cycle life, the level of discharge depth, and the level of predicted temperature, and then taking the highest level among the level of predicted cycle life, the level of predicted discharge depth, and the level of predicted temperature as the hidden danger level.

9. The method for monitoring the energy storage status of an energy storage power station according to claim 1, characterized in that: Also includes: S6. Generate countermeasures and suggestions based on the prediction results; The countermeasures and suggestions include monitoring suggestions and maintenance suggestions.

Images