Cooperative optimization method and device of energy storage system, electronic equipment and storage medium

Abstract

The present disclosure discloses a collaborative optimization method and device of an energy storage system, an electronic device and a storage medium. According to the present application, a training library is constructed by integrating system historical operation data and environmental parameters, the influence of environmental changes on operation efficiency is dynamically predicted by means of a prediction model, the compressor and the expander are collaboratively and rollingly optimized under the safety constraint of the equipment, and the model is corrected in real time to adapt to the variable working condition demand. Therefore, the technical problems of heat exchanger efficiency reduction, difficulty in maintaining the best work capacity of the expander, frequent adjustment of the system and efficiency loss under variable working conditions caused by independent optimization of the compressor and the expander, inability to adapt to dynamic changes of environmental parameters and coupling constraints of equipment performance in the existing operation optimization method can be solved. The technical effects of improving the system operation efficiency under variable working conditions, adapting to environmental changes in different seasons and day-night temperature differences, reducing invalid adjustment of the system, breaking through the coupling constraints of equipment performance, and ensuring stable and efficient operation of the system under full working conditions are achieved.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a collaborative optimization method and apparatus, electronic device and storage medium for energy storage systems. Background Technology

[0002] Compressed air energy storage, as a core technology in the field of large-scale physical energy storage, is widely used in renewable energy consumption and grid peak shaving scenarios. Related technologies utilize the coordinated operation of multi-stage compression / expansion systems and thermal management devices to construct a complete energy conversion system encompassing air compression, storage, reheating, and expansion power generation.

[0003] Existing operational optimization methods are prone to heat exchanger efficiency decline in high-temperature summer environments and struggle to maintain optimal expander performance in low-temperature winter conditions. Furthermore, the diurnal temperature variations in spring and autumn cause frequent adjustments to the system's operating status. Therefore, the existing independent optimization modes for the compressor and expander cannot overcome the constraints of performance coupling between the devices, resulting in efficiency losses under varying operating conditions. Summary of the Invention

[0004] This disclosure provides a collaborative optimization method, apparatus, electronic device, and storage medium for an energy storage system.

[0005] According to a first aspect of this disclosure, a collaborative optimization method for an energy storage system is provided, comprising:

[0006] Collect and process historical operating data and corresponding environmental parameter data of the compressed air energy storage system to construct a parameter library for model training;

[0007] A prediction model is established and trained based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters based on the input of dynamically changing environmental parameters.

[0008] Based on the real-time acquired environmental parameter prediction data for future periods, and based on the prediction model, the operating parameters of the compressor and expander are optimized online in a rolling manner while meeting the constraints of safe operation of the equipment, so as to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize system efficiency.

[0009] The optimal operating pressure ratio and expansion ratio are output as control commands to the system controller to coordinate and control the operating status of the compressor and expander;

[0010] The system acquires real-time operating data and performs online correction on the prediction model based on the deviation between the actual operating data and the predicted values.

[0011] Optionally, the collection and processing of historical operational data and environmental parameter data includes:

[0012] For missing data points, interpolation is used to fill them in, and abnormal data that exceeds the preset physical reasonable range is removed.

[0013] The environmental parameter data and system operation data are aligned according to a unified time base.

[0014] Optionally, the prediction model is a time-series prediction model based on neural networks, which processes time-dependent input data through a gating mechanism.

[0015] Optionally, the rolling online optimization of the operating parameters of the compressor and expander includes:

[0016] The solution is obtained by using an evolutionary algorithm or a swarm intelligence optimization algorithm, where the decision variables of the algorithm include the operating parameters of the multi-stage compressor and the multi-stage expander.

[0017] Optionally, the coordinated control of the operating states of the compressor and expander includes:

[0018] The operating pressure ratio can be adjusted by regulating the compressor's speed or the opening of its inlet guide vanes;

[0019] The expansion ratio is adjusted by regulating the speed of the expander.

[0020] Optionally, the online calibration of the prediction model includes:

[0021] When the deviation between the model's predicted efficiency and the actual operating efficiency continues to exceed a preset threshold, the model update process is triggered.

[0022] The prediction model is periodically retrained and its parameters are updated based on a training set composed of newly collected operational data and historical data.

[0023] According to a second aspect of this disclosure, a collaborative optimization device for an energy storage system is provided, comprising:

[0024] The construction unit is also used to collect and process historical operating data and corresponding environmental parameter data of the compressed air energy storage system, and to build a parameter library for model training.

[0025] The establishment unit is also used to establish and train a prediction model based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters based on the input of dynamically changing environmental parameters.

[0026] The optimization unit is also used to perform rolling online optimization of the operating parameters of the compressor and expander based on the prediction model, according to the real-time acquired environmental parameter prediction data for future periods, under the condition of meeting the equipment safety operation constraints, so as to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize the system efficiency.

[0027] The control unit is also used to output the optimal operating pressure ratio and expansion ratio as control commands to the system controller in order to coordinate and control the operating status of the compressor and expander;

[0028] The correction unit is also used to acquire the actual operating data of the system in real time and to correct the prediction model online based on the deviation between the actual operating data and the predicted value.

[0029] Optionally, the building unit is further configured to:

[0030] For missing data points, interpolation is used to fill them in, and abnormal data that exceeds the preset physical reasonable range is removed.

[0031] The environmental parameter data and system operation data are aligned according to a unified time base.

[0032] Optionally, the prediction model established by the establishment unit is a time-series prediction model based on neural networks, and the model processes time-dependent input data through a gating mechanism.

[0033] Optionally, the optimization unit is further configured to:

[0034] The solution is obtained by using an evolutionary algorithm or a swarm intelligence optimization algorithm, where the decision variables of the algorithm include the operating parameters of the multi-stage compressor and the multi-stage expander.

[0035] Optionally, the control unit is further configured to:

[0036] The operating pressure ratio can be adjusted by regulating the compressor's speed or the opening of its inlet guide vanes;

[0037] The expansion ratio is adjusted by regulating the speed of the expander.

[0038] Optionally, the correction unit is further configured to:

[0039] When the deviation between the model's predicted efficiency and the actual operating efficiency continues to exceed a preset threshold, the model update process is triggered.

[0040] The prediction model is periodically retrained and its parameters are updated based on a training set composed of newly collected operational data and historical data.

[0041] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0042] At least one processor; and

[0043] A memory communicatively connected to the at least one processor; wherein,

[0044] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0045] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0046] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0047] The collaborative optimization method, apparatus, electronic equipment, and storage medium for energy storage systems disclosed herein, through the integration of historical system operating data and environmental parameters to construct a training library, dynamically predicts the impact of environmental changes on operating efficiency using a predictive model, and performs collaborative rolling online optimization of the compressor and expander under equipment safety constraints, and corrects the model through real-time data to adapt to varying operating conditions, can solve the technical problems in existing operation optimization methods caused by independent optimization of the compressor and expander, inability to adapt to dynamic changes in environmental parameters and equipment performance coupling constraints, resulting in decreased heat exchanger efficiency, difficulty in maintaining the optimal work capacity of the expander, frequent system adjustments, and efficiency losses under varying operating conditions. This achieves the technical effects of improving system operating efficiency under varying operating conditions, adapting to environmental changes in different seasons and diurnal temperature ranges, reducing ineffective system adjustments, and thus overcoming equipment performance coupling limitations and ensuring stable and efficient system operation under all operating conditions.

[0048] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0049] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0050] Figure 1 A flowchart illustrating a collaborative optimization method for an energy storage system provided in an embodiment of this disclosure;

[0051] Figure 2 A schematic diagram of the structure of a collaborative optimization device for an energy storage system provided in an embodiment of this disclosure;

[0052] Figure 3 A schematic block diagram of an example electronic device provided for embodiments of this disclosure. Detailed Implementation

[0053] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0054] The following description, with reference to the accompanying drawings, outlines a collaborative optimization method, apparatus, electronic device, and storage medium for energy storage systems according to embodiments of the present disclosure.

[0055] Figure 1 This is a flowchart illustrating a collaborative optimization method for an energy storage system provided in an embodiment of this disclosure.

[0056] like Figure 1 As shown, the method includes the following steps:

[0057] Step 101: Collect and process the historical operating data and corresponding environmental parameter data of the compressed air energy storage system, and construct a parameter library for model training;

[0058] Historical operational data must cover key state parameters during the operation of the compressed air energy storage system, including system power generation, power consumption, compressor stage outlet pressure and temperature, expander stage inlet pressure and temperature, air storage device pressure, and heat storage device temperature. This data can be systematically extracted from the system's control system historical database or long-term operational records to ensure a complete reflection of the system's performance under different operating conditions. Environmental parameter data focuses on meteorological data related to the system's location, including ambient temperature, ambient pressure, and ambient humidity. This data can be obtained through the system's supporting environmental sensors, a professional meteorological monitoring platform, or long-term accumulated historical meteorological records, ensuring that the data accurately corresponds to the system's operating time and spatial scenarios.

[0059] After data acquisition, systematic processing is required. Appropriate methods should be used to fill in any temporary data gaps that may occur during the acquisition process, ensuring data integrity. Simultaneously, abnormal data exceeding physically reasonable limits should be strictly identified and removed to prevent invalid data from interfering with model training. Subsequently, data alignment processing is performed. Using the compressed air energy storage system's operating clock as a unified benchmark, the time scales of historical operating data and environmental parameter data are adjusted to be consistent, establishing a one-to-one correspondence so that each set of system operating data can accurately match the environmental parameter data for the same time period.

[0060] After a series of processing steps, including data collection, completion, anomaly removal, and alignment, the resulting complete dataset becomes the parameter library for model training. This parameter library can comprehensively and accurately present the system's operating rules under different environmental conditions, providing a solid data foundation for the subsequent model learning of the complex relationship between environmental parameters and system operating efficiency, and ensuring the accuracy and reliability of model training.

[0061] Step 102: Establish and train a prediction model based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters based on the input of dynamically changing environmental parameters.

[0062] The parameter library integrates historical operating data and corresponding environmental parameter data, covering system operating characteristics under different working conditions and environmental conditions. This provides a rich and comprehensive sample foundation for model learning, ensuring that the model can fully explore the complex mapping relationship between changes in environmental parameters and the expected operating efficiency of the system. The input of the prediction model focuses on the dynamic changes in environmental parameters, with core parameters including ambient temperature, ambient pressure, and ambient humidity. These parameters directly affect the heat exchange effect, working fluid state, and equipment operating characteristics of the compressed air energy storage system, and are key factors causing fluctuations in system operating efficiency.

[0063] By learning the dynamic changes of these environmental parameters over time and their synergistic effects with the system's set operating parameters, the model can accurately identify the changing trends of system efficiency under different environmental scenarios. The model's output is explicitly the expected operating efficiency of the system under the set operating parameters. This efficiency index intuitively reflects the system's energy conversion level when a specific combination of operating parameters is adapted to current and future environmental conditions, providing clear guidance for the subsequent optimization of operating parameters.

[0064] During model training, the training, validation, and test sets should be appropriately divided based on the parameter library data. Adaptive training strategies should be employed to adjust the model structure and parameters. Continuous iteration should reduce the deviation between predicted and actual values. Simultaneously, an effective regularization mechanism should be introduced to prevent overfitting and ensure the model possesses good generalization ability. The trained prediction model can quickly respond to dynamic changes in environmental parameters. Even in scenarios with drastic environmental fluctuations, it can accurately output reliable expected operating efficiency for the set operating parameters, providing real-time and accurate prediction data for subsequent rolling online optimization. This ensures that the optimization algorithm can solve for the optimal operating parameters based on true and effective efficiency prediction results, thus laying a solid foundation for the efficient operation of the system.

[0065] Step 103: Based on the real-time acquired environmental parameter prediction data for future periods, and based on the prediction model, under the condition of meeting the equipment safety operation constraints, perform rolling online optimization of the operating parameters of the compressor and expander to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize system efficiency.

[0066] First, real-time forecast data of environmental parameters for future periods is required. This data, encompassing ambient temperature, pressure, and humidity, can be obtained through professional meteorological forecasting systems or supporting environmental monitoring equipment. This ensures the data reflects environmental trends over a future period, providing a realistic input basis for optimization calculations. Subsequently, the trained prediction model is used as an efficiency evaluation tool. This model does not directly output optimal operating parameters but, based on the input environmental parameter forecast data, predicts the expected system operating efficiency for any set of candidate operating parameters (such as pressure ratio and expansion ratio), providing an evaluation basis for the objective function value of the optimization algorithm. The optimization process must strictly adhere to equipment safety constraints. These constraints are based on the equipment's physical characteristics and safety standards, fundamentally ensuring the stable operation of the compressor and expander and preventing equipment failure or performance degradation due to abnormal parameters.

[0067] The optimization employs a rolling online optimization mode, proceeding cyclically according to a preset control cycle. Each cycle restarts the optimization calculation by incorporating the latest updated environmental parameter prediction data and the real-time system status. Through continuous iterative adjustment of operating parameters, the uncertainty caused by dynamic changes in environmental parameters is effectively mitigated, ensuring that the optimization results are always highly adapted to the current operating conditions. The core objective of the optimization is to achieve optimal system efficiency. Under the constraint of safe equipment operation, the optimization algorithm iteratively searches for combinations of operating parameters for the compressor and expander. In each iteration, the optimization algorithm generates a set of candidate parameters and inputs them into the prediction model for evaluation. The model returns the expected efficiency under this parameter combination, and the optimization algorithm continuously adjusts its search direction accordingly, ultimately selecting the operating parameter combination that maximizes the overall efficiency of the system in the future, namely the optimal operating pressure ratio of the compressor and the optimal operating expansion ratio of the expander.

[0068] Step 104: Output the optimal operating pressure ratio and expansion ratio as control commands to the system controller to coordinate and control the operating status of the compressor and expander;

[0069] The optimal operating pressure ratio and optimal operating expansion ratio obtained through rolling online optimization are core control parameters that adapt to environmental parameters in the future forecast period and enable the system to achieve optimal efficiency. Their accuracy and adaptability directly determine the actual operating effect of the system.

[0070] These optimal parameters are treated as explicit control commands and precisely output to the system controller. As the core hub connecting upper-level optimization decisions and lower-level equipment operation, the system controller must respond quickly and analyze these control commands, combining the real-time operating status of the compressor and expander to formulate a coordinated control strategy. The core of coordinated control lies in ensuring a high degree of matching between the operating states of the compressor and expander, avoiding operational imbalances and efficiency losses caused by independent operation. By adjusting the compressor speed or guide vane opening, the actual operating pressure ratio of the compressor accurately tracks the optimal operating pressure ratio. Simultaneously, by adjusting the expander speed, the actual operating expansion ratio of the expander strictly matches the optimal operating expansion ratio, achieving dynamic coordination of their operating rhythms and performance parameters.

[0071] During control execution, the system controller continuously receives operational feedback signals from the compressor and expander, monitors the deviation between the actual and optimal values ​​of the pressure ratio and expansion ratio in real time, and fine-tunes control commands accordingly. This ensures that the equipment's operating state consistently fluctuates around the optimal operating point, preventing deviations from the optimal efficiency range due to minor changes in the external environment or equipment operational deviations. This collaborative control mode not only fully leverages the effectiveness of optimized parameters but also ensures the operational safety and coordination of the compressor and expander, reduces equipment wear, extends service life, and promotes the system to maintain a consistently efficient and stable operating state under dynamically changing environmental conditions, effectively achieving the optimization goals.

[0072] Step 105: Acquire the actual operating data of the system in real time, and perform online correction of the prediction model based on the deviation between the actual operating data and the predicted value.

[0073] During system operation, continuous real-time acquisition of actual operating data is required. This data encompasses two core dimensions: first, the actual operating efficiency of the system, which directly reflects the degree of agreement between the model's predictions and actual operating conditions, serving as the core basis for deviation calculation; and second, environmental parameters for the corresponding time period, including ambient temperature, pressure, and humidity, ensuring that deviation analysis can accurately pinpoint the correlation between the environment and system operation. Data acquisition relies on the system's supporting sensor monitoring equipment and underlying control system to maintain continuity and timeliness, avoiding inaccurate correction judgments due to data delays or missing data. After data collection, the actual operating efficiency of the system is directly compared with the previously output expected operating efficiency of the prediction model, accurately calculating the deviation value between the two. The magnitude of the deviation value directly reflects the current prediction accuracy level of the model.

[0074] To avoid unnecessary frequent corrections caused by a single abnormal data point or short-term operating condition fluctuations, a reasonable deviation threshold needs to be set. The online model correction procedure should only be triggered when the calculated deviation value reaches or exceeds this threshold, ensuring the targetedness and effectiveness of the correction action. During the correction process, the latest collected actual operating data and corresponding environmental parameters are integrated as new valid samples into the data foundation. An adapted model update algorithm is then used to adjust the parameters and retrain the prediction model, correcting the prediction biases that previously existed in the model.

[0075] This online correction mechanism enables the model to adapt in a timely manner to changes brought about by the slow aging of system components, the performance degradation of equipment, and long-term changes in environmental patterns. It continuously optimizes the model's adaptability to complex operating conditions, ensuring that the model maintains high prediction accuracy throughout long-term operation. This provides stable and reliable efficiency prediction support for subsequent rolling online optimization, ensuring that the compressed air energy storage system can achieve efficient and stable operation based on accurate predictions throughout its entire life cycle.

[0076] In some embodiments, the collection and processing of historical operational data and environmental parameter data includes:

[0077] For missing data points, interpolation is used to fill them in, and abnormal data that exceeds the preset physical reasonable range is removed.

[0078] The environmental parameter data and system operation data are aligned according to a unified time base.

[0079] During data acquisition, some data points may be missing due to factors such as temporary sensor malfunctions and signal transmission interference. Such omissions disrupt the continuity and integrity of the data, thus affecting the model's ability to learn data patterns. Therefore, interpolation is used to fill in the missing data points, supplementing missing information by reasonably fitting data trends to ensure the coherence of the data sequence. Simultaneously, the acquired data may contain abnormal data exceeding the preset physical reasonable range. This type of data is often caused by abnormal factors such as equipment malfunctions and extreme interference, and lacks reference value for actual operating conditions. Retaining it would seriously mislead the model's training direction. Therefore, it is necessary to strictly identify and remove such abnormal data to ensure that every set of data in the dataset conforms to actual physical operating patterns.

[0080] Environmental parameter data and system operation data may originate from different monitoring devices, and the time recording benchmarks of each device differ. Direct use of these data would result in inaccurate mapping to the same operating period, losing the meaningful correlation between the two. Therefore, both types of data need to be adjusted to a unified time benchmark, ensuring that each piece of environmental parameter data accurately matches the system operation data at the same moment, establishing a one-to-one correspondence. This guarantees that subsequent model training can accurately capture the intrinsic relationship between environmental changes and system operating status. Through the above data processing steps, problems such as missing, abnormal, and asynchronous data that may exist in the original data are effectively solved, significantly improving the reliability and effectiveness of the data. This lays a solid foundation for building a high-quality parameter library, ensuring that subsequent model training can be based on real and valid data, thereby improving the accuracy and stability of model predictions.

[0081] In some embodiments, the prediction model is a neural network-based time-series prediction model, which processes time-dependent input data through a gating mechanism.

[0082] The prediction model employs a neural network-based time-series prediction model. Its core advantage lies in its ability to accurately capture the time dependencies in the input data and adapt to the dynamic changes in environmental parameters and system operating data. Based on a neural network architecture, the time-series prediction model achieves efficient processing of time-series data through a unique gating mechanism. This mechanism intelligently regulates the transmission and retention of information, avoiding the information loss or gradient vanishing problems that occur in traditional neural networks when processing long-sequence data.

[0083] Gating mechanisms can flexibly control the degree of retention of historical information, filtering out past data features valuable for current predictions, while precisely controlling the fusion ratio of new input information. This ensures that the model can respond promptly to dynamic fluctuations in environmental parameters and fully learn the intrinsic relationship between environmental conditions and system operating efficiency at different times. Because environmental parameters such as temperature, pressure, and humidity exhibit continuous changes, and the evolution of system operating states has a significant temporal continuity, the time dependence of input data directly affects the accuracy of prediction results. Gating mechanisms can effectively address this key issue, enabling the model to deeply mine long-term patterns and short-term fluctuations in the data.

[0084] During training, the model uses a gating structure to process input data from multiple time periods in an orderly manner, gradually building a complex mapping relationship between temporal changes in environmental parameters and the expected operating efficiency of the system. Whether it's a sudden change in short-term environmental parameters or a shift in long-term environmental patterns, the model can effectively capture these changes and transform them into accurate efficiency predictions. This neural network-based time-series prediction model, thanks to the unique role of its gating mechanism, significantly improves its ability to process time-dependent data, ensuring that the output of the expected system operating efficiency has high timeliness and reliability. This provides solid predictive support for subsequent rolling online optimization, ensuring that the optimization algorithm can solve for the optimal operating parameters based on accurate efficiency predictions, and driving the compressed air energy storage system to maintain high efficiency under dynamic environmental conditions.

[0085] In some embodiments, the rolling online optimization of the operating parameters of the compressor and expander includes:

[0086] The solution is obtained by using an evolutionary algorithm or a swarm intelligence optimization algorithm, where the decision variables of the algorithm include the operating parameters of the multi-stage compressor and the multi-stage expander.

[0087] When performing rolling online optimization of the operating parameters of compressors and expanders, evolutionary algorithms or swarm intelligence optimization algorithms are selected as the core solution tools. These algorithms possess powerful global search capabilities and adaptability to nonlinear problems, effectively addressing challenges such as complex constraints involving multivariate coupling in the optimization of operating parameters for multi-stage compressors and expanders. Evolutionary algorithms simulate natural selection and genetic mutation mechanisms, continuously filtering for the optimal solution through population iteration. Swarm intelligence optimization algorithms, on the other hand, draw inspiration from the cooperative behavior of biological groups, relying on information exchange and collaborative search among individuals to achieve a globally optimal solution. Both types of algorithms can escape the trap of local optima and accurately discover parameter combinations that meet the goal of maximizing system efficiency.

[0088] The decision variables in the optimization process comprehensively cover the key operating parameters of multi-stage compressors and multi-stage expanders, including the speed and guide vane opening of each stage of compressor and the speed of each stage of expander. These parameters directly determine the working state and energy conversion efficiency of each stage of equipment and are the core factors affecting the overall system performance. Simultaneous optimization of the operating parameters of multi-stage equipment as unified decision variables can overcome the limitations of independent adjustment of single-stage equipment, achieve operational coordination between each stage of compressor and expander, and avoid overall system efficiency losses caused by local parameter optimization.

[0089] Under the rolling online optimization framework, the algorithm re-solves the decision variables based on the latest environmental parameter prediction data and the real-time system status in each control cycle. By continuously iterating and updating the parameter combination, it ensures that the operating parameters of each level of equipment are always highly adapted to the dynamically changing environmental conditions and system requirements, thereby driving the overall system efficiency to remain at the optimal level while meeting the constraints of safe equipment operation.

[0090] In some embodiments, the coordinated control of the operating states of the compressor and expander includes:

[0091] The operating pressure ratio can be adjusted by regulating the compressor's speed or the opening of its inlet guide vanes;

[0092] The expansion ratio is adjusted by regulating the speed of the expander.

[0093] The core of coordinating the operation of the compressor and expander lies in precisely adjusting key operating parameters to ensure that the actual operating pressure ratio and expansion ratio of both are strictly matched to the optimal values ​​obtained through rolling online optimization, thus ensuring that the system is always in a state of efficient and coordinated operation. For the compressor, the operating pressure ratio is a key indicator that determines its compression efficiency and energy consumption. By adjusting the compressor speed, the compression intensity and rate of air in the compression chamber can be directly changed. Changes in speed will synchronously affect the degree to which air is compressed, thereby achieving precise control of the operating pressure ratio. When it is necessary to increase the pressure ratio, the speed can be appropriately increased; when it is necessary to decrease the pressure ratio, the speed can be decreased accordingly.

[0094] Adjusting the opening of the compressor inlet guide vanes is also an effective way to adjust the operating pressure ratio. The change in the opening of the guide vanes will affect the airflow and airflow angle entering the compressor, causing an adaptive change in the compression process inside the compressor, thereby indirectly adjusting the operating pressure ratio. This method is especially suitable for scenarios that require rapid response to pressure ratio adjustment needs or small-range fine adjustments. The two adjustment methods can be flexibly selected or combined according to the actual working conditions to ensure that the compressor operating pressure ratio stably tracks the optimal set value.

[0095] For expanders, the operating expansion ratio directly affects their work capacity and power generation efficiency. By adjusting the expander's speed, the degree of air expansion and the work process within the expander can be directly controlled. Appropriate speed adjustment ensures that the airflow expansion process inside the expander closely matches the optimal expansion ratio requirement. When the speed is adjusted to the suitable range, the air can fully expand and perform work within the expander, maximizing the conversion of internal energy into mechanical energy and then into electrical energy. Furthermore, speed adjustment is characterized by rapid response and high control precision, accurately tracking changes in the optimal operating expansion ratio. Throughout the coordinated control process, the system controller receives real-time operational feedback signals from the compressor and expander, continuously monitoring the deviation between the actual pressure-to-expansion ratio and the optimal value. It promptly adjusts the adjustment range and pace to ensure coordinated operation between the two, preventing a decrease in overall system efficiency due to deviations in parameters of a single device. Ultimately, this achieves efficient coordinated operation of the compressor and expander, driving the compressed air energy storage system to maintain stable and high energy conversion efficiency under dynamic environmental conditions.

[0096] In some embodiments, the online calibration of the prediction model includes:

[0097] When the deviation between the model's predicted efficiency and the actual operating efficiency continues to exceed a preset threshold, the model update process is triggered.

[0098] The prediction model is periodically retrained and its parameters are updated based on a training set composed of newly collected operational data and historical data.

[0099] The core of online calibration of the prediction model is to ensure that the model maintains accurate prediction capabilities over the long term through a dynamic update mechanism. Its implementation must strictly adhere to the dual logic of deviation triggering and periodic retraining. During real-time system operation, the prediction efficiency output by the prediction model and the actual operating efficiency of the system are continuously monitored. By calculating the deviation between the two in real time, the current prediction accuracy of the model is determined. The preset threshold is set based on the acceptable range of prediction error for engineering applications. Its function is to filter out temporary deviations caused by single abnormal data points or short-term fluctuations in operating conditions, preventing the model from being frequently triggered for updates by invalid interference. Only when the deviation is continuously detected to exceed the preset threshold does it indicate that the model's prediction capability can no longer adapt to the current system state or environmental changes. At this point, the model update process is immediately triggered to ensure the necessity and specificity of the calibration action.

[0100] The core of the model update process is to build a high-quality hybrid training set. The newly collected operational data includes the actual operating efficiency of the system in the latest time period, the corresponding environmental parameters, and the system operating status data. This data can accurately reflect the latest changes such as the current performance status of system components and changes in environmental modes. On the other hand, historical data covers the long-term operating patterns under different working conditions and environmental conditions. By organically mixing the two types of data, it is ensured that the training set can cover new operational features while retaining the model's learning results on long-term patterns, thus avoiding the loss of historical information due to training with only new data.

[0101] Based on this, the prediction model is retrained according to a preset period. The period setting needs to balance model update efficiency and system operation stability. By continuously iterating and adjusting the model's network weights and internal parameters, the model can adapt to performance changes caused by factors such as the slow aging of system components and long-term changes in environmental conditions, and continuously correct prediction biases. This online correction method, which combines a bias triggering mechanism with periodic retraining, ensures both the timeliness and effectiveness of model updates, and guarantees the model's generalization ability and long-term stability through a hybrid training set. This allows the prediction model to continuously output accurate and reliable expected operating efficiency throughout its entire lifecycle, providing solid support for the rolling online optimization of compressed air energy storage systems and ensuring that the system always operates within its optimal efficiency range.

[0102] Corresponding to the aforementioned collaborative optimization method for energy storage systems, this invention also proposes a collaborative optimization device for energy storage systems. Since the device embodiments of this invention correspond to the aforementioned method embodiments, details not disclosed in the device embodiments can be referred to the aforementioned method embodiments, and will not be repeated here.

[0103] Figure 2 This is a schematic diagram of the structure of a collaborative optimization device for an energy storage system provided in an embodiment of this disclosure, as shown below. Figure 2 As shown, it includes:

[0104] The construction unit 21 is also used to collect and process the historical operating data and corresponding environmental parameter data of the compressed air energy storage system, and to build a parameter library for model training.

[0105] The establishment unit 22 is also used to establish and train a prediction model based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters according to the input of dynamically changing environmental parameters.

[0106] The optimization unit 23 is also used to perform rolling online optimization of the operating parameters of the compressor and expander based on the prediction model according to the real-time acquired environmental parameter prediction data for future periods, under the condition of meeting the equipment safety operation constraints, so as to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize the system efficiency.

[0107] The control unit 24 is also used to output the optimal operating pressure ratio and expansion ratio as control commands to the system controller in order to coordinate and control the operating status of the compressor and expander;

[0108] The correction unit 25 is also used to acquire the actual operating data of the system in real time, and to perform online correction of the prediction model based on the deviation between the actual operating data and the predicted value.

[0109] Furthermore, in one possible implementation of this disclosure embodiment, the construction unit 21 is further configured to:

[0110] For missing data points, interpolation is used to fill them in, and abnormal data that exceeds the preset physical reasonable range is removed.

[0111] The environmental parameter data and system operation data are aligned according to a unified time base.

[0112] Furthermore, in one possible implementation of this disclosure embodiment, the prediction model established by the establishment unit 22 is a time-series prediction model based on a neural network, and the model processes time-dependent input data through a gating mechanism.

[0113] Furthermore, in one possible implementation of this disclosure, the optimization unit 23 is further configured to:

[0114] The solution is obtained by using an evolutionary algorithm or a swarm intelligence optimization algorithm, where the decision variables of the algorithm include the operating parameters of the multi-stage compressor and the multi-stage expander.

[0115] Furthermore, in one possible implementation of this disclosure embodiment, the control unit 24 is further configured to:

[0116] The operating pressure ratio can be adjusted by regulating the compressor's speed or the opening of its inlet guide vanes;

[0117] The expansion ratio is adjusted by regulating the speed of the expander.

[0118] Furthermore, in one possible implementation of this disclosure, the correction unit 25 is further configured to:

[0119] When the deviation between the model's predicted efficiency and the actual operating efficiency continues to exceed a preset threshold, the model update process is triggered.

[0120] The prediction model is periodically retrained and its parameters are updated based on a training set composed of newly collected operational data and historical data.

[0121] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of the embodiments of this disclosure, and the principle is the same. Therefore, the embodiments of this disclosure are not limited thereto.

[0122] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0123] Figure 3 A schematic block diagram of an example electronic device 400 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0124] like Figure 3 As shown, device 400 includes a computing unit 401, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 402 or a computer program loaded from storage unit 408 into RAM (Random Access Memory) 403. RAM 403 may also store various programs and data required for the operation of device 400. The computing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. I / O (Input / Output) interface 405 is also connected to bus 404.

[0125] Multiple components in device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of monitors, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0126] The computing unit 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 401 performs the various methods and processes described above, such as the co-optimization method for energy storage systems. For example, in some embodiments, the co-optimization method for energy storage systems can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 408. In some embodiments, part or all of the computer program can be loaded and / or installed on device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by the computing unit 401, one or more steps of the methods described above can be performed. Alternatively, in other embodiments, computing unit 401 may be configured to perform the aforementioned collaborative optimization method for the energy storage system by any other suitable means (e.g., by means of firmware).

[0127] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0128] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0129] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0130] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0131] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0132] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0133] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0134] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0135] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A collaborative optimization method for an energy storage system, characterized in that, include: Collect and process historical operating data and corresponding environmental parameter data of compressed air energy storage systems, and build a parameter library for model training; A prediction model is established and trained based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters based on the dynamically changing input of environmental parameters. The prediction model is a time-series prediction model based on neural networks, and the model processes time-dependent input data through a gating mechanism. Based on the real-time acquired environmental parameter prediction data for future periods, and based on the prediction model, the operating parameters of the compressor and expander are optimized online in a rolling manner while meeting the constraints of safe operation of the equipment, so as to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize system efficiency. The optimal operating pressure ratio and expansion ratio are output as control commands to the system controller to coordinate and control the operating status of the compressor and expander; The system acquires real-time actual operating data and performs online correction on the prediction model based on the deviation between the actual operating data and the predicted values. The rolling online optimization of the operating parameters of the compressor and expander includes: The solution is obtained by using an evolutionary algorithm or a swarm intelligence optimization algorithm, where the decision variables of the algorithm include the operating parameters of the multi-stage compressor and the multi-stage expander.

2. The method according to claim 1, characterized in that, The collection and processing of historical operational data and environmental parameter data includes: For missing data points, interpolation is used to fill them in, and abnormal data that exceeds the preset physical reasonable range is removed. The environmental parameter data and system operation data are aligned according to a unified time base.

3. The method according to claim 1, characterized in that, The coordinated control of the operating status of the compressor and expander includes: The operating pressure ratio can be adjusted by regulating the compressor's speed or the opening of its inlet guide vanes; The expansion ratio is adjusted by regulating the speed of the expander.

4. The method according to claim 1, characterized in that, The online calibration of the prediction model includes: When the deviation between the model's predicted efficiency and the actual operating efficiency continues to exceed a preset threshold, the model update process is triggered. The prediction model is periodically retrained and its parameters are updated based on a training set composed of newly collected operational data and historical data.

5. A collaborative optimization device for an energy storage system, characterized in that, include: The building unit is also used to collect and process historical operating data and corresponding environmental parameter data of compressed air energy storage systems, and to build a parameter library for model training. The establishment unit is also used to establish and train a prediction model based on the parameter library. The prediction model is used to output the expected operating efficiency of the system under set operating parameters based on the input of dynamically changing environmental parameters. The optimization unit is also used to perform rolling online optimization of the operating parameters of the compressor and expander based on the prediction model, according to the real-time acquired environmental parameter prediction data for future periods, under the condition of meeting the equipment safety operation constraints, so as to obtain the compressor operating pressure ratio and expander operating expansion ratio that optimize the system efficiency. The control unit is also used to output the optimal operating pressure ratio and expansion ratio as control commands to the system controller in order to coordinate and control the operating status of the compressor and expander; The correction unit is also used to acquire the actual operating data of the system in real time and to correct the prediction model online based on the deviation between the actual operating data and the predicted value.

6. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-4.

7. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-4.

8. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-4.

Images