Operation analysis method of novel energy storage system based on big data processing

Abstract

The invention relates to the technical field of energy storage systems, in particular to an operation analysis method of a novel energy storage system based on big data processing. The method comprises the following steps: performing feature extraction on data in a fusion database by using a big data processing platform to obtain feature sets under different time nodes; and inputting the feature set of the plurality of preset time nodes into a neural network, outputting a corresponding estimated value of the battery health state, inputting the estimated values of the plurality of battery health states into a prediction model, and outputting the remaining service life of the battery. Inputting a constraint condition constructed at least based on the estimated value of the state of health of the battery and the predicted value of the remaining service life of the battery and a target function into an optimization solver, and outputting a plan of charging and discharging power in a preset time period; and based on the planned operation, counting the standard deviation of the voltages of the plurality of battery modules after the operation for a preset time to determine the operation state, and performing adjustment based on the operation state. The operation analysis efficiency of the energy storage system is improved.

Description

Technical Field

[0001] This invention relates to the field of energy storage system technology, and in particular to a novel operation analysis method for energy storage systems based on big data processing. Background Technology

[0002] With the large-scale grid connection of renewable energy and the popularization of electric vehicles, energy storage systems are playing an increasingly crucial role in grid peak shaving, frequency regulation, renewable energy consumption, and user-side energy management. However, the operation and management of existing energy storage systems face many challenges. For example, existing systems mainly rely on local data from battery management systems, converter systems, and data acquisition and monitoring control systems, resulting in a single data dimension and a lack of deep integration with external environmental data. Traditional analysis is mostly based on threshold alarms and simple statistics, which cannot accurately predict the system's health status and remaining lifespan, nor can it identify complex performance degradation patterns. Control strategies are mostly preset fixed strategies, lacking dynamic optimization capabilities based on real-time data and future predictions, making it difficult to maximize economic benefits or minimize system losses. Operation and maintenance are usually carried out after a failure occurs or performance degrades significantly, lacking predictive maintenance capabilities, leading to high operation and maintenance costs and low system availability. Therefore, there is an urgent need for a method that can deeply integrate multi-source data and utilize advanced big data and AI technologies to achieve intelligent analysis, prediction, and optimization of the entire lifecycle of energy storage systems.

[0003] Chinese Patent Publication No. CN112688378A discloses an energy storage system operation control method, device, and energy storage system, including: acquiring the current output power and charge / discharge cycle count of all energy storage units; obtaining the priority value of energy storage output of each energy storage unit based on the current output power and charge / discharge cycle count of all energy storage units; and controlling energy storage output based on the priority value of energy storage output of each energy storage unit.

[0004] It is evident that the existing technology suffers from the following problems: fragmented data analysis and low level of intelligent decision-making, resulting in low operational analysis efficiency for new energy storage systems. Summary of the Invention

[0005] To address this, the present invention provides a novel energy storage system operation analysis method based on big data processing, which overcomes the problems of fragmented data analysis and low level of intelligent decision-making in existing technologies, resulting in low efficiency of operation analysis for novel energy storage systems.

[0006] To achieve the above objectives, this invention provides a novel operation analysis method for energy storage systems based on big data processing, comprising: The system collects multi-data information, including internal operational data, external environmental data, and historical data, through a data interface using a multi-sensor array. The collected multi-data elements are merged based on a unified timestamp to build a fused database; A big data processing platform is built based on the fused database, and the big data processing platform is used to extract features from the data in the fused database to obtain feature sets at different time points. The feature sets of multiple preset time points are respectively input into the neural network model trained based on the historical data, and the estimated value of the battery health status corresponding to the preset time point is output. The estimated values ​​at multiple preset time points are input into the prediction model, and the predicted value of the remaining battery life is output. The constraints and objective function, which are constructed based at least on the estimated value and the predicted value, are input into the optimization solver, and the optimal solution of the charging and discharging power within a preset time period is output, wherein the constraints include at least the maximum charging and discharging power. The optimal solution is executed using an energy storage converter, and the standard deviation of the voltage of multiple battery modules is calculated after a preset execution time. The parameters of the PID controller inside the energy storage converter are adjustable. The operating status of the novel energy storage system is determined based on the standard deviation, and the maximum charge and discharge power in the constraints is adjusted based on the operating status. The proportional gain of the PID controller is then adjusted based on the operating status after adjusting the maximum charge and discharge power in the constraints.

[0007] Furthermore, determining the operating status of the novel energy storage system based on the standard deviation includes: if the standard deviation of the voltages of multiple battery modules is less than or equal to a preset standard deviation, the operating status of the novel energy storage system is determined to be qualified; if the standard deviation of the voltages of multiple battery modules is greater than the preset standard deviation, the operating status of the novel energy storage system is determined to be unqualified. The step of adjusting the maximum charge and discharge power in the constraints based on the operating status includes: if the operating status is determined to be unqualified, drawing a heat map based on the acquired battery module voltage and temperature; if the heat map has a gradient, adjusting the maximum charge and discharge power in the constraints based on the ratio of the standard deviation of multiple battery module voltages to a preset standard deviation.

[0008] Furthermore, adjusting the maximum charge / discharge power in the constraint based on the ratio of the standard deviation of the voltages of multiple battery modules to the preset standard deviation includes: reducing the maximum charge / discharge power in the constraint based on the ratio of the standard deviation of the voltages of multiple battery modules to the preset standard deviation, and the reduction in the maximum charge / discharge power is proportional to the ratio.

[0009] Furthermore, the method also includes: if the operating state after adjusting the maximum charging and discharging power is unqualified, then obtain the standard deviation of the voltage of multiple battery modules after multiple historical execution preset times; draw the execution number-standard deviation curve and calculate the integral of the curve; if the integral of the curve is greater than the preset integral, then adjust the optimization period of the initial parameters of the neural network based on the ratio of the integral of the curve to the preset integral.

[0010] Furthermore, adjusting the optimization period of the initial parameters of the neural network based on the ratio of the integral of the curve to the preset integral includes: reducing the optimization period of the initial parameters of the neural network based on the ratio of the integral of the curve to the preset integral, and the reduction in the optimization period is proportional to the ratio.

[0011] Furthermore, the method further includes: if the operating state after adjusting the optimization cycle is unqualified, then the optimization cycle of the initial parameters of the neural network is repeatedly adjusted at least once until the number of adjustments is less than a preset number and the operating state is qualified, or the number of adjustments is equal to the preset number and the adjustment is stopped; if the operating state is unqualified after the adjustment is stopped, then the optimal solution for multiple time nodes within a preset time period and the power feedback value output by the energy storage converter corresponding to the same time node are obtained; the time node-optimal solution curve and the time node-power feedback value curve are plotted respectively, and the integrals of the time node-optimal solution curve and the time node-power feedback value curve are calculated to obtain the first integral and the second integral; the absolute value of the difference between the first integral and the second integral is calculated; if the absolute value is greater than a preset absolute value, the proportional gain of the controller in the energy storage converter is adjusted based on the ratio of the absolute value to the preset absolute value.

[0012] Furthermore, adjusting the proportional gain of the controller in the energy storage converter based on the ratio of the absolute value to the preset absolute value includes: increasing the proportional gain of the PID controller in the energy storage converter based on the ratio of the absolute value to the preset absolute value, and the increase in proportional gain is proportional to the ratio.

[0013] Furthermore, the method also includes: increasing the derivative gain of the PID controller in the energy storage converter based on the ratio of a preset proportional gain to the proportional gain, wherein the increase in derivative gain is inversely proportional to the ratio.

[0014] Furthermore, the method also includes: if the operating state after adjusting the differential gain is unqualified, then calculate the average value of the battery health state estimate corresponding to multiple preset time nodes; if the average value is greater than the preset average value, then adjust the inspection cycle of the multi-sensor array based on the difference between the average value and the preset average value.

[0015] Furthermore, adjusting the verification cycle of the multi-sensor array based on the difference between the average value and the preset average value includes: reducing the verification cycle of the multi-sensor array based on the difference between the average value and the preset average value, and the reduction in the verification cycle is proportional to the difference.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention fuses multi-source data acquired by a multi-sensor array to construct a fusion database, and utilizes a big data processing platform built on this fusion database to extract features from the data in the fusion database, obtaining feature sets at different time points, thus accurately extracting data features. Furthermore, by inputting the feature sets of multiple preset time points into a neural network, it outputs corresponding estimates of battery health status, accurately grasping the battery health status at different preset time points. The estimated values ​​of multiple battery health statuses are input into a prediction model to output the remaining battery lifespan. Constraints and objective functions are constructed based at least on the estimated battery health status and the predicted remaining battery lifespan. These constraints and objective functions are input into an optimization solver, which outputs a schedule of charging and discharging power within a preset time period, effectively operating the novel energy storage system. The operating status is determined by statistically analyzing the standard deviation of the voltages of multiple battery modules after a preset execution time, and adjustments are made based on the operating status. This invention, by integrating multi-source data and utilizing advanced big data and artificial intelligence technologies, achieves panoramic perception and intelligent diagnosis of the energy storage system, thereby improving the efficiency of energy storage system operation analysis.

[0017] Furthermore, when the operating status of the novel energy storage system is unqualified, the present invention determines whether to adjust the maximum charge and discharge power in the constraints based on the heat map drawn by obtaining the battery module voltage and temperature. This allows for a more effective determination of the cause of the unqualified operation based on the heat map, thereby enabling more accurate adjustment of the operating parameters based on the cause and further improving the efficiency of the energy storage system's operation analysis.

[0018] Furthermore, this invention adjusts the maximum charging and discharging power in the constraint conditions based on the ratio of the standard deviation of multiple battery module voltages to the preset standard deviation. This allows for more accurate adjustment of the maximum charging and discharging power in the constraint conditions, thereby enabling timely reduction of system heat generation and voltage deviation when the cooling capacity decreases along the path due to unreasonable duct design or insufficient airflow speed. This further improves the operational analysis efficiency of the energy storage system.

[0019] Furthermore, this invention determines the optimization period of the neural network's initial parameters based on the integral of the execution time period-standard deviation curve. This allows for a more accurate determination of the reasons for unqualified operating conditions, thereby enabling more effective adjustment of operating parameters and further improving the operational analysis efficiency of the energy storage system.

[0020] Furthermore, this invention adjusts the optimization period of the initial parameters of the neural network based on the ratio of the integral of the curve to the preset integral. This allows for a more accurate adjustment of the optimization period of the initial parameters of the neural network, enabling the model to make decisions based on the actual capacity of the weakest module in the current system. This results in more accurate model decision data and further improves the operational analysis efficiency of the energy storage system.

[0021] Furthermore, when the operating state is unqualified after repeatedly adjusting the initial parameters of the neural network for several optimization cycles, the present invention determines whether to adjust the proportional gain of the controller in the energy storage converter based on the absolute value of the difference between the first integral and the second integral. This allows for adjustments based on more accurate reasons, thereby further improving the efficiency of the energy storage system's operation analysis.

[0022] Furthermore, this invention adjusts the proportional gain of the PID controller in the energy storage converter based on the ratio of the absolute value to the preset absolute value. This allows for more accurate adjustment of the proportional gain of the PID controller in the energy storage converter, enabling the energy storage converter to respond more promptly to the execution command of the optimal solution after outputting the optimal solution for charging and discharging power, thereby further improving the operational analysis efficiency of the energy storage system.

[0023] Furthermore, this invention adjusts the derivative gain of the PID controller in the energy storage converter based on the ratio of a preset proportional gain to the proportional gain. This can more effectively avoid overshoot and oscillation caused by excessive adjustment after improving the response speed, thereby further improving the operational analysis efficiency of the energy storage system.

[0024] Furthermore, when the operating state is unqualified after adjusting the differential gain, the present invention determines whether to adjust the verification cycle of the multi-sensor array based on the average value of the battery health state estimates corresponding to multiple preset time nodes. This allows for more effective adjustment of the corresponding parameters based on the reasons for the unqualified operating state, thereby further improving the operational analysis efficiency of the energy storage system.

[0025] Furthermore, the present invention adjusts the verification cycle of the multi-sensor array based on the difference between the average value and the preset average value, which can more accurately adjust the verification cycle of the multi-sensor array, thereby more accurately determining the battery health status estimate, and further improving the operation analysis efficiency of the energy storage system. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the operation analysis system of a novel energy storage system based on big data processing, as described in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the steps of the operation analysis method for a novel energy storage system based on big data processing, as described in an embodiment of the present invention. Figure 3This is a flowchart illustrating the steps of determining the standard deviation of multiple battery module voltages relative to a preset standard deviation in an embodiment of the present invention. Figure 4 This is a flowchart illustrating the steps for determining the running state based on the optimized cycle of the initial parameters of the neural network in an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0028] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0029] It should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0030] Please see Figure 1 As shown, it is a schematic diagram of the operation analysis system of the novel energy storage system based on big data processing according to an embodiment of the present invention.

[0031] The system includes an acquisition unit, a fusion unit, a feature extraction unit, an estimated value output unit, a predicted value output unit, an optimal solution output unit, a calculation unit, and an analysis unit.

[0032] The acquisition unit collects multi-data information obtained by the multi-sensor array through a data interface, including internal operating data, external environmental data, and historical data; The fusion unit is connected to the acquisition unit and is used to fuse the acquired multi-data points based on a unified timestamp to construct a fusion database. The feature extraction unit is connected to the fusion unit and is used to build a big data processing platform based on the fusion database, and to use the big data processing platform to extract features from the data in the fusion database to obtain feature sets at different time points. The estimated value output unit is connected to the feature extraction unit, and is used to input the feature sets of multiple preset time nodes into the neural network model trained based on the historical data, and output the estimated value of the battery health status corresponding to the preset time node. The predicted value output unit is connected to the estimated value output unit, and is used to input the estimated values ​​of multiple preset time points into the prediction model and output the predicted value of the remaining battery life. The optimal solution output unit is connected to the estimated value output unit and the predicted value output unit respectively. It is used to input the constraints and objective function constructed based at least on the estimated value and the predicted value into the optimization solver and output the optimal solution of the charging and discharging power within a preset time period. The constraints include at least the maximum charging and discharging power. The calculation unit is connected to the optimal solution output unit, which is used to execute the optimal solution using the energy storage converter and calculate the standard deviation of the voltage of multiple battery modules after a preset execution time. The parameters of the PID controller inside the energy storage converter are adjustable. The analysis unit is connected to the calculation unit and is used to determine the operating status of the new energy storage system based on the standard deviation, adjust the maximum charge and discharge power in the constraints based on the operating status, and adjust the proportional gain of the PID controller based on the operating status after adjusting the maximum charge and discharge power in the constraints.

[0033] Specifically, internal operating data includes real-time data from the battery management system, inverter, and local monitoring system, such as battery module voltage, current, temperature, battery status, power, and energy throughput. External environmental data includes meteorological data, real-time electricity prices, forecasted electricity prices, and real-time electricity revenue.

[0034] Specifically, a big data platform is built using a distributed computing framework for the storage, computation, and management of massive amounts of time-series data, such as feature extraction. This process is existing technology and will not be elaborated here.

[0035] Specifically, a multidimensional feature sequence within a time window, such as data from the past 30 charge-discharge cycles, is input into a trained neural network model, such as a long short-term memory network model. The model learns the long-term dependencies in the sequence through its internal memory units and finally outputs an accurate estimate of the current battery health status.

[0036] Specifically, a convolutional neural network is used to perform deep feature extraction on the historical battery health status estimate, and the extracted feature sequence is input into a neural network model, such as a long short-term memory network model, to learn the long-term time dependence of capacity decay and the overall degradation trajectory. The model predicts the estimated value of future battery health status, and by extrapolating to the end-of-life threshold, the predicted value of the remaining battery life can be obtained. This process is existing technology and will not be described in detail here.

[0037] Specifically, the optimization solver is an open-source mathematical optimization solver, and the constraints are constructed based on the maximum available capacity, maximum charging power, and maximum discharging power. Wherein, the maximum usable capacity Q_usable(t) = SOH(t) * Q_rated, where SOH(t) is the estimated value of the battery health state at time node t; Maximum charging power and maximum discharging power are determined based on battery health state estimates and lookup table functions; The objective function is Maximize(Σ [electricity revenue - degradation cost]), where the degradation cost Cost_deg(t) = f(SOH, RUL, DoD, C-rate) * E_throughput(t), where RUL is the predicted remaining battery life, DoD is the depth of discharge, and E_throughput(t) is the battery's throughput energy at time node t.

[0038] Specifically, the purpose of calculating the standard deviation of the voltages of multiple battery modules after the preset execution time is to check the voltage differences of each battery module under static or low-current conditions after the preset execution time of the strategy, so as to determine the system balance state.

[0039] Please see Figure 2 The diagram shown is a flowchart illustrating the steps of an operation analysis method for a novel energy storage system based on big data processing, according to an embodiment of the present invention.

[0040] S1, through the acquisition unit and data interface, acquires multi-data information obtained by the multi-sensor array, including internal operating data, external environmental data and historical data; S2, the collected multi-data elements are fused based on a unified timestamp by a fusion unit connected to the acquisition unit to construct a fusion database; S3, a big data processing platform is constructed based on the fusion database by a feature extraction unit connected to the fusion unit, and the big data processing platform is used to extract features from the data in the fusion database to obtain feature sets at different time points; S4, the feature sets of multiple preset time nodes are respectively input into the neural network model trained based on the historical data through the estimation output unit connected to the feature extraction unit, and the estimated value of the battery health status corresponding to the preset time node is output. S5, the estimated values ​​of multiple preset time nodes are input into the prediction model through the prediction value output unit connected to the estimated value output unit, and the predicted value of the remaining battery life is output. S6, the optimal solution output unit, which is connected to the estimated value output unit and the predicted value output unit respectively, inputs the constraints and objective function constructed based on the estimated value and the predicted value into the optimization solver, and outputs the optimal solution of the charging and discharging power within a preset time period, wherein the constraints include at least the maximum charging and discharging power; S7, the optimal solution is executed by the energy storage converter through the calculation unit connected to the optimal solution output unit, and the standard deviation of the voltage of multiple battery modules is calculated after a preset execution time. The parameters of the PID controller inside the energy storage converter are adjustable. S8, the analysis unit connected to the computing unit determines the operating status of the new energy storage system based on the standard deviation, adjusts the maximum charging and discharging power in the constraints based on the operating status, and adjusts the proportional gain of the PID controller based on the operating status after adjusting the maximum charging and discharging power in the constraints.

[0041] Please see Figure 3 The diagram shows a flowchart illustrating the steps of determining the operating status of a novel energy storage system based on the standard deviation of multiple battery module voltages in an embodiment of the present invention. The method for determining the operating status of the novel energy storage system based on the standard deviation in this embodiment includes: if the standard deviation of the multiple battery module voltages is less than or equal to the preset standard deviation, the operating status of the novel energy storage system is determined to be qualified; if the standard deviation of the multiple battery module voltages is greater than the preset standard deviation, the operating status of the novel energy storage system is determined to be unqualified.

[0042] Specifically, taking the improvement of the economic efficiency and safety of new energy storage systems throughout their entire life cycle as an example, and based on the chemical characteristics, thermodynamic limits and power market regulation requirements of battery cells, and combined with key information such as system attenuation trajectory, failure modes and revenue fluctuations obtained from the statistical analysis of massive historical operating data, subsequent corresponding preset or critical parameter values ​​are set.

[0043] Specifically, taking low-current charging and discharging as an example, with a preset standard deviation L0 = 20 mV, the comparison process between the standard deviation L of multiple battery module voltages and the preset standard deviation L0 is as follows: If the standard deviation L of the voltage of multiple battery modules is less than or equal to the preset standard deviation L0, the operating status of the new energy storage system is determined to be qualified. If the standard deviation L of the voltages of multiple battery modules is greater than the preset standard deviation L0, the operating status of the new energy storage system is determined to be unqualified.

[0044] Specifically, if the heatmap plotted based on the acquired battery module voltage and temperature exhibits a gradient—that is, a "hot spot" (highest temperature) and a "voltage anomaly point" (highest voltage during charging / lowest voltage during discharging)—it indicates that in the air-cooling system, the temperature is low at one end of the fan and high at the far end. This suggests an unreasonable airflow design or insufficient airflow speed, leading to a decrease in cooling capacity along the cooling path. In this case, it is necessary to adjust the maximum charge / discharge power in the constraints. This can immediately reduce the current flowing through the faulty module, thereby reducing its heat generation and voltage deviation. The maximum charge / discharge power in the constraints is adjusted based on the ratio of the standard deviation of multiple battery module voltages to a preset standard deviation. The preset ratio P0 = 1.2 of the standard deviation of multiple battery module voltages to the preset standard deviation is as follows: If the ratio P of the standard deviation of multiple battery module voltages to the preset standard deviation is less than or equal to the preset ratio P0, then the maximum charging power will be adjusted to 0.91 times the original maximum charging power, and the maximum discharging power will be adjusted to 0.93 times the original maximum discharging power. If the ratio P of the standard deviation of multiple battery module voltages to the preset standard deviation is greater than the preset ratio P0, then the maximum charging power will be adjusted to 0.85 times the original maximum charging power, and the maximum discharging power will be adjusted to 0.87 times the original maximum discharging power.

[0045] Specifically, the operating state after adjusting the maximum charging and discharging power is redefined. If the operating state is unqualified, the standard deviation of the voltage of multiple battery modules after multiple historical execution preset times is obtained, and the execution number-standard deviation curve is plotted. The integral of the curve is calculated. If the integral of the curve is greater than the preset integral, it indicates that the neural network is still using the initial parameters of the system and has not performed real-time updates and calibration. The model does not know the actual bearing capacity of the weakest module in the current system, thus causing the operating state to be unqualified. Therefore, the optimization cycle of the initial parameters of the neural network is adjusted based on the ratio of the integral of the curve to the preset integral. The preset ratio of the integral of the curve to the preset integral is Q0=1.5. The comparison process between the ratio Q of the integral of the curve to the preset integral and the preset ratio Q0 is as follows: If the ratio Q of the integral of the curve to the preset integral is less than or equal to the preset ratio Q0, the optimization period of the initial parameters of the neural network is adjusted to 0.86 times the original optimization period. The unit of the optimization period is hours, and the adjusted values ​​are all rounded up. If the ratio Q of the integral of the curve to the preset integral is greater than the preset ratio Q0, the optimization period of the initial parameters of the neural network will be adjusted to 0.72 times the original optimization period. The unit of the optimization period is hours, and the adjusted values ​​are all rounded up.

[0046] Please see Figure 4 As shown, it is a flowchart of the steps for determining the running state based on the optimized period of the initial parameters of the neural network in an embodiment of the present invention.

[0047] Specifically, after re-determining the optimization period of the initial parameters of the neural network, the operating state is adjusted. If the operating state is unsatisfactory, the optimization period of the initial parameters of the neural network is adjusted at least once until the number of adjustments is less than a preset number and the operating state is satisfactory, or the number of adjustments is equal to the preset number. If the operating state is still unsatisfactory after stopping the adjustment, the optimal solutions for multiple time nodes within a preset time period and the power feedback value output by the energy storage converter corresponding to the same time node are obtained. The time node-optimal solution curve and the time node-power feedback value curve are plotted respectively, and the integral of the time node-optimal solution curve is calculated. The first integral is obtained, and the integral of the time node-power feedback value curve is calculated to obtain the second integral. The absolute value of the difference between the first integral and the second integral is calculated. If the absolute value is greater than the preset absolute value, it indicates that the energy storage converter takes a long time to respond to the execution command of the optimal solution after the optimal solution of the output charging and discharging power, i.e., response delay. Therefore, the response speed needs to be improved. The proportional gain of the PID controller in the energy storage converter is adjusted based on the ratio of the absolute value to the preset absolute value. The preset ratio R0 is 1.2. The specific comparison process between the absolute value and the preset absolute value R and the preset ratio R0 is as follows: If the ratio R of the absolute value to the preset absolute value is less than or equal to the preset ratio R0, then the proportional gain of the controller in the energy storage converter will be adjusted to 1.2 times the original proportional gain. If the ratio R of the absolute value to the preset absolute value is greater than the preset ratio R0, then the proportional gain of the controller in the energy storage converter will be adjusted to 1.47 times the original proportional gain.

[0048] Specifically, after improving the response speed, to prevent overshoot and oscillation caused by excessive adjustment, the derivative gain of the PID controller in the energy storage converter is adjusted based on the ratio of the preset proportional gain to the preset proportional gain. The preset ratio of the preset proportional gain to the preset proportional gain is T0 = 1.35. The comparison process between the preset ratio T0 and the preset proportional gain ratio T is as follows: If the ratio T of the preset proportional gain to the original proportional gain is less than or equal to the preset ratio T0, then the derivative gain of the PID controller in the energy storage converter will be adjusted to 1.45 times the original derivative gain. If the ratio T of the preset proportional gain to the original proportional gain is greater than the preset ratio T0, then the derivative gain of the PID controller in the energy storage converter will be adjusted to 1.19 times the original derivative gain.

[0049] Specifically, the operating status is re-checked after adjusting the differential gain. If the operating status is still unqualified, the average value of the battery health status estimate corresponding to multiple preset time points is calculated. If the average value is greater than the preset average value, it indicates that the battery health status estimate is changing slowly and does not conform to the normal change pattern, meaning that there is a problem with the multi-sensor array in acquiring data. Therefore, the verification cycle of the multi-sensor array is adjusted based on the difference between the average value and the preset average value. The preset difference between the average value and the preset average value is U0 = 1.5%. The comparison process between the difference U and the preset average value is as follows: If the difference U between the average value and the preset average value is less than or equal to the preset difference U0, the verification cycle of the multi-sensor array will be adjusted to 0.94 times the original verification cycle. The unit of the verification cycle is hours, and the adjusted values ​​are all rounded up. If the difference U between the average value and the preset average value is greater than the preset difference U0, the verification cycle of the multi-sensor array will be adjusted to 0.86 times the original verification cycle. The unit of the verification cycle is hours, and the adjusted values ​​are all rounded up.

[0050] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0051] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A novel operation analysis method for energy storage systems based on big data processing, characterized in that, include: The system collects multi-data information, including internal operational data, external environmental data, and historical data, through a data interface using a multi-sensor array. The collected multi-data elements are merged based on a unified timestamp to build a fused database; A big data processing platform is built based on the fused database, and the big data processing platform is used to extract features from the data in the fused database to obtain feature sets at different time points. The feature sets of multiple preset time points are respectively input into the neural network model trained based on the historical data, and the estimated value of the battery health status corresponding to the preset time point is output. The estimated values ​​at multiple preset time points are input into the prediction model, and the predicted value of the remaining battery life is output. The constraints and objective function, which are constructed based at least on the estimated value and the predicted value, are input into the optimization solver, and the optimal solution of the charging and discharging power within a preset time period is output, wherein the constraints include at least the maximum charging and discharging power. The optimal solution is executed using an energy storage converter, and the standard deviation of the voltage of multiple battery modules is calculated after a preset execution time. The parameters of the PID controller inside the energy storage converter are adjustable. The operating status of the novel energy storage system is determined based on the standard deviation, and the maximum charge and discharge power in the constraints is adjusted based on the operating status. The proportional gain of the PID controller is then adjusted based on the operating status after adjusting the maximum charge and discharge power in the constraints.

2. The operation analysis method for a novel energy storage system based on big data processing according to claim 1, characterized in that, Determining the operating status of the novel energy storage system based on the standard deviation includes: If the standard deviation of the voltage of multiple battery modules is less than or equal to the preset standard deviation, the operating status of the new energy storage system is determined to be qualified. If the standard deviation of the voltage of multiple battery modules is greater than the preset standard deviation, the operating status of the new energy storage system is determined to be unqualified. The adjustment of the maximum charge / discharge power in the constraints based on the operating state includes: If the operating condition is determined to be unqualified, a heat map is drawn based on the acquired battery module voltage and temperature; If the heatmap has a gradient, the maximum charge / discharge power in the constraints is adjusted based on the ratio of the standard deviation of the voltages of multiple battery modules to the preset standard deviation.

3. The operation analysis method for a novel energy storage system based on big data processing according to claim 2, characterized in that, The adjustment of the maximum charge / discharge power in the constraint condition based on the ratio of the standard deviation of multiple battery module voltages to a preset standard deviation includes: The maximum charge and discharge power in the constraint condition is reduced by the ratio of the standard deviation of multiple battery module voltages to the preset standard deviation, and the reduction of the maximum charge and discharge power is proportional to the ratio.

4. The operation analysis method for a novel energy storage system based on big data processing according to claim 3, characterized in that, The method further includes: If the operating status is not qualified after adjusting the maximum charging and discharging power, then obtain the standard deviation of the voltage of multiple battery modules after multiple historical execution preset times. Plot the number of executions versus standard deviation curve and calculate the integral of the curve; If the integral of the curve is greater than the preset integral, the optimization period of the initial parameters of the neural network is adjusted based on the ratio of the integral of the curve to the preset integral.

5. The operation analysis method for a novel energy storage system based on big data processing according to claim 4, characterized in that, The optimization period for adjusting the initial parameters of the neural network based on the ratio of the curve integral to a preset integral includes: The ratio of the integral of the curve to the preset integral reduces the optimization period of the initial parameters of the neural network, and the reduction in the optimization period is proportional to the ratio.

6. The operation analysis method for a novel energy storage system based on big data processing according to claim 5, characterized in that, The method further includes: If the running status is not satisfactory after adjusting the optimization cycle, the optimization cycle of the initial parameters of the neural network is adjusted at least once, until the number of adjustments is less than the preset number and the running status is satisfactory, or the number of adjustments is equal to the preset number and the adjustment stops. If the operating status is not up to standard after the adjustment is stopped, the optimal solution for multiple time nodes within a preset time period and the power feedback value output by the energy storage converter corresponding to the same time node are obtained. Plot the time node-optimal solution curve and the time node-power feedback value curve respectively, and calculate the integral of the time node-optimal solution curve and the integral of the time node-power feedback value curve to obtain the first integral and the second integral; Calculate the absolute value of the difference between the first integral and the second integral; If the absolute value is greater than the preset absolute value, the proportional gain of the controller in the energy storage converter is adjusted based on the ratio of the absolute value to the preset absolute value.

7. The operation analysis method for a novel energy storage system based on big data processing according to claim 6, characterized in that, The method of adjusting the proportional gain of the controller in the energy storage converter based on the ratio of the absolute value to a preset absolute value includes: The proportional gain of the PID controller in the energy storage converter is increased based on the ratio of the absolute value to the preset absolute value, and the increase in proportional gain is proportional to the ratio.

8. The operation analysis method for a novel energy storage system based on big data processing according to claim 7, characterized in that, The method further includes: The derivative gain of the PID controller in the energy storage converter is increased based on the ratio of the preset proportional gain to the proportional gain, and the increase in derivative gain is inversely proportional to the ratio.

9. The operation analysis method for a novel energy storage system based on big data processing according to claim 8, characterized in that, The method further includes: If the operating state is not up to standard after adjusting the differential gain, calculate the average value of the battery health state estimate corresponding to multiple preset time nodes. If the average value is greater than the preset average value, the verification cycle of the multi-sensor array will be adjusted based on the difference between the average value and the preset average value.

10. The operation analysis method for a novel energy storage system based on big data processing according to claim 9, characterized in that, The adjustment of the multi-sensor array verification cycle based on the difference between the average value and the preset average value includes: The verification cycle of the multi-sensor array is reduced based on the difference between the average value and the preset average value, and the reduction in the verification cycle is proportional to the difference.

Images