Monitoring control method and device based on compressed air energy storage power station

Abstract

The invention provides a monitoring control method and device based on a compressed air energy storage power station. Acquiring multi-dimensional monitoring data of the compressed air energy storage power station in a current time period and multi-dimensional monitoring data of the compressed air energy storage power station in a reference time period; wherein the multi-dimensional monitoring data comprises thermodynamic state data and air component data; the reference time period at least comprises a compression time period and an expansion time period which are connected; the method comprises the following steps: determining a theoretical reference value for a current time period according to multi-dimensional monitoring data of a reference time period by utilizing a preset online simulation model; performing deviation comparison according to the theoretical reference value of the current time period and the multi-dimensional monitoring data of the current time period to obtain a deviation comparison result, and extracting time domain evolution characteristics of the current time period according to the deviation comparison result; determining a fault type according to the time domain evolution characteristics of the current time period; and performing corresponding operation control on the compressed air energy storage power station according to the fault type. Therefore, the monitoring dimension is expanded, and the operation and maintenance cost is reduced.

Description

Technical Field

[0001] This specification belongs to the field of compressed air energy storage technology, and in particular relates to a monitoring and control method and device based on a compressed air energy storage power station. Background Technology

[0002] Existing compressed air energy storage power station monitoring systems generally adopt an operation mode that focuses on monitoring key component status parameters and coordinates with maintenance personnel for real-time control. However, this traditional control method has significant drawbacks, including limited system monitoring range, delayed control response time, and high overall operation and maintenance costs.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This specification provides a monitoring and control method and device for compressed air energy storage power stations, which solves the problems of limited monitoring range, reliance on manual control leading to delayed response time, and high overall operation and maintenance costs in existing compressed air energy storage power station monitoring systems.

[0005] This specification provides a monitoring and control method for a compressed air energy storage power station, the method including:

[0006] Acquire multidimensional monitoring data of the compressed air energy storage power station for the current time period, as well as multidimensional monitoring data for a reference time period; wherein, the multidimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period;

[0007] By utilizing a preset online simulation model and based on multidimensional monitoring data from the reference time period, a theoretical reference value for the current time period is determined.

[0008] Based on the theoretical reference value and the multidimensional monitoring data of the current time period, a deviation comparison is performed to obtain the deviation comparison result. Based on the deviation comparison result, the temporal evolution characteristics of the current time period are extracted.

[0009] Based on the time-domain evolution characteristics of the current time period, the fault type is determined; and the compressed air energy storage power station is operated accordingly based on the fault type.

[0010] In one embodiment, determining the theoretical reference value for the current time period by utilizing a preset online simulation model and based on multidimensional monitoring data from the reference time period includes:

[0011] Based on the multidimensional monitoring data of the reference time period, the boundary conditions of the preset online simulation model are set using the thermodynamic state data of the reference time period to obtain the initialized model;

[0012] Input the air composition data for the reference time period into the initialized model;

[0013] The fouling thermal resistance parameters in the initial model were updated using particulate matter content data; the thermophysical parameters in the initial model were updated using moisture content data, resulting in the updated model.

[0014] Based on the updated model, the thermodynamic equations are solved to calculate the theoretical reference value for the current time period. In one embodiment, the deviation comparison is performed between the theoretical reference value and the multidimensional monitoring data for the current time period to obtain the deviation comparison result. Based on the deviation comparison result, the temporal evolution characteristics of the current time period are extracted, including:

[0015] Based on the theoretical reference value for the current time period and the multidimensional monitoring data for the current time period, the deviation is compared to determine the corresponding residual sequence;

[0016] The residual sequence is subjected to time-domain feature extraction to obtain the time-domain evolution features of the current time period; wherein the time-domain evolution features include at least the rate of change feature value and the trend pattern feature.

[0017] In one embodiment, determining the fault type based on the time-domain evolution characteristics of the current time period includes:

[0018] When the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a monotonically divergent state, the fault type is determined to be a gradual performance degradation fault caused by the cumulative effect of air components.

[0019] When the rate of change characteristic value is lower than the preset mutation threshold and the trend pattern characteristic shows a non-monotonic divergent state, the fault type is determined to be a transient disturbance fault and filtered out.

[0020] When the characteristic value of the rate of change is not lower than the preset mutation threshold, the fault type is determined to be a sudden mechanical fault originating from the equipment itself.

[0021] In one embodiment, the step of performing corresponding operational control of the compressed air energy storage power station according to the fault type includes:

[0022] When the fault type is the gradual performance degradation fault, adjust the first speed setting value of the low-temperature circulating pump installed on the connecting pipeline between the low-temperature heat storage tank and the compression end heat exchanger, and / or adjust the second speed setting value of the high-temperature circulating pump installed on the connecting pipeline between the high-temperature heat storage tank and the expansion end heat exchanger, and monitor the heat exchange efficiency of the high-temperature circulating pump and / or the low-temperature circulating pump until the heat exchange efficiency reaches the preset deviation range.

[0023] In one embodiment, the step of performing corresponding operational control of the compressed air energy storage power station according to the fault type includes:

[0024] When the fault type is the sudden mechanical fault, the first switch valve located at the salt cave inlet section and the second switch valve located at the salt cave outlet section are controlled to perform a rapid closing action, and the power input of the compressor unit or expander unit is cut off simultaneously.

[0025] In one embodiment, the method further includes:

[0026] Using a parallel time-series feature extraction module of a preset thermodynamic digital twin evolution model, the corresponding time-series feature vector sequence is determined based on multi-dimensional monitoring data of the reference time period;

[0027] Using the physical constraint decoding module of the preset thermodynamic digital twin evolution model, under the constraint of thermodynamic energy conservation boundary conditions, the theoretical reference value of the current time period is determined according to the time-series feature vector sequence;

[0028] The preset thermodynamic digital twin evolution model is a deep neural network model based on a heterogeneous dual-stream architecture; the parallel temporal feature extraction module is a module based on a bidirectional long short-term memory network structure with independent dual-channel encoding; and the physical constraint decoding module is a module based on a multilayer perceptron and energy manifold correction structure.

[0029] This specification provides a monitoring and control device for a compressed air energy storage power station, comprising:

[0030] The data acquisition module is used to acquire multi-dimensional monitoring data of the compressed air energy storage power station for the current time period, as well as multi-dimensional monitoring data for a reference time period; wherein, the multi-dimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period;

[0031] The reference value determination module is used to determine the theoretical reference value for the current time period by using a preset online simulation model and based on the multidimensional monitoring data of the reference time period.

[0032] The feature extraction module is used to compare the deviations between the theoretical reference values ​​and the multidimensional monitoring data of the current time period, obtain the deviation comparison results, and extract the temporal evolution features of the current time period based on the deviation comparison results.

[0033] The operation control module is used to determine the fault type based on the time-domain evolution characteristics of the current time period; and to perform corresponding operation control on the compressed air energy storage power station according to the fault type.

[0034] This specification also provides an electronic device, including a processor and a memory for storing processor-executable instructions, wherein the processor, when executing the instructions, implements a monitoring and control method for a compressed air energy storage power station.

[0035] This specification also provides a computer-readable storage medium storing computer instructions that, when executed, implement a monitoring and control method for a compressed air energy storage power station.

[0036] Based on the monitoring and control method for a compressed air energy storage power station provided in this specification, multi-dimensional monitoring data of the compressed air energy storage power station for the current time period and multi-dimensional monitoring data for a reference time period are acquired. The multi-dimensional monitoring data includes thermodynamic state data and air composition data. The air composition data includes at least particulate matter content data and moisture content data. The reference time period includes at least one consecutive compression time period and expansion time period. Using a preset online simulation model, a theoretical reference value for the current time period is determined based on the multi-dimensional monitoring data of the reference time period. A deviation comparison is performed between the theoretical reference value and the multi-dimensional monitoring data for the current time period to obtain the deviation comparison result. Based on the deviation comparison result, the time-domain evolution characteristics of the current time period are extracted. Based on the time-domain evolution characteristics of the current time period, the fault type is determined. And based on the fault type, corresponding operation control is performed on the compressed air energy storage power station. In this way, by acquiring air composition data including particulate matter content and moisture content, and incorporating this air composition data into multidimensional monitoring data, the monitoring dimensions of the system are effectively expanded. Simultaneously, this method utilizes a pre-set online simulation model to accurately determine the theoretical reference value for the current time period based on multidimensional monitoring data containing at least one consecutive compression and expansion time period. Furthermore, it extracts time-domain evolution features based on deviation comparison results to automatically determine the fault type and implement corresponding operational control. This achieves timely fault identification and precise fault management, significantly solving the technical shortcomings of delayed control response time and high overall operation and maintenance costs. Attached Figure Description

[0037] To more clearly illustrate the embodiments of this specification, the accompanying drawings used in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart illustrating a monitoring and control method for a compressed air energy storage power station, provided in one embodiment of this specification.

[0039] Figure 2 This is a schematic diagram of the electronic device structure provided in one embodiment of this specification;

[0040] Figure 3 This is a schematic diagram of the structural composition of a monitoring and control device for a compressed air energy storage power station, provided in one embodiment of this specification.

[0041] Figure 4 This is an overall connection diagram provided in one embodiment of this specification;

[0042] Figure 5 This is a schematic diagram illustrating the working principle of one embodiment of this specification;

[0043] Figure 6 This is a schematic diagram illustrating the steps of an intelligent control method provided in one embodiment of this specification.

[0044] Legend:

[0045] 1. Salt cavern; 11. First switching valve; 12. Second switching valve; 13. Salt cavern inlet air temperature and pressure sensor; 14. Salt cavern inlet air component analyzer; 2. Compressor unit; 101. Compressor end inlet air filter; 102. Compressor end gas flow meter; 103. Compressor unit inlet air component analyzer; 21. Compressor unit inlet air temperature and pressure sensor; 22. Compressor unit speed and vibration sensor; 3. Expander unit; 31. Expander unit speed and vibration sensor; 4. Cryogenic storage tank; 41. Cryogenic circulating pump; 42. First regulating valve; 43. First flow meter; 44. 45. Low-temperature temperature and pressure sensor; 5. Low-temperature liquid level gauge; 6. High-temperature heat storage tank; 7. High-temperature circulating pump; 8. Second regulating valve; 9. Second flow meter; 10. High-temperature temperature and pressure sensor; 11. High-temperature liquid level gauge; 22. Compression end heat exchanger; 33. Compression end heat exchanger inlet and outlet temperature and pressure sensor; 44. Expansion end heat exchanger; 55. Expansion end gas flow meter; 6. Expansion unit inlet air component detector; 76. Expansion end heat exchanger inlet and outlet temperature and pressure sensor; 8. Gas-liquid separator; 9. Gas-liquid separator level gauge; 10. Silencer. Detailed Implementation

[0046] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0047] See Figure 1 As shown in the embodiments of this specification, a monitoring and control method based on a compressed air energy storage power station is provided, wherein the method is specifically applied to the server side. In specific implementation, the method may include the following:

[0048] S101: Acquire multi-dimensional monitoring data of the compressed air energy storage power station for the current time period, and multi-dimensional monitoring data for a reference time period; wherein, the multi-dimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period;

[0049] S102: By utilizing a preset online simulation model, a theoretical reference value for the current time period is determined based on the multi-dimensional monitoring data of the reference time period;

[0050] S103: Based on the theoretical reference value and the multidimensional monitoring data of the current time period, perform deviation comparison to obtain the deviation comparison result, and extract the time-domain evolution characteristics of the current time period based on the deviation comparison result.

[0051] S104: Determine the fault type based on the time-domain evolution characteristics of the current time period; and perform corresponding operation control on the compressed air energy storage power station according to the fault type.

[0052] Among them, the aforementioned multidimensional monitoring data can refer to the heterogeneous data set collected in real time during the operation of the compressed air energy storage power station. This breaks through the limitations of traditional methods that only monitor physical field parameters, and integrates air composition data reflecting the composition of the working fluid with thermodynamic state data reflecting energy conversion in a multidimensional way.

[0053] The aforementioned thermodynamic state data can refer to macroscopic physical quantity data obtained by temperature and pressure sensors and flow meters deployed at key nodes such as power plant compressors, expanders, heat exchangers and salt caverns. These data mainly cover parameters such as temperature, pressure and flow rate, and are used to characterize the work state, fluid flow characteristics and energy transfer efficiency at the current moment.

[0054] The aforementioned air composition data can refer to quantitative indicators that reflect the purity and composition of the air working fluid, collected by dedicated detection instruments. It clearly includes particulate matter content data (affecting heat exchanger fouling and flow resistance) and water content data (affecting fluid specific heat capacity and compressibility factor), and is a key input variable for the simulation model to perform parameter correction and evolutionary deduction.

[0055] The aforementioned compression period can be the process of using electricity to drive a compressor to inject air into a salt cavern for energy storage, and the aforementioned expansion period can be the process of releasing high-pressure air to drive an expander to generate electricity.

[0056] The aforementioned theoretical reference value can refer to the expected ideal operating value for the current time period generated by using a preset online simulation model and performing evolutionary deduction based on multi-dimensional monitoring data of a reference time period (i.e., historical period).

[0057] The aforementioned deviation comparison result refers to the real-time residual sequence generated by performing difference calculations between the actual multidimensional monitoring data collected in the current time period and the derived theoretical reference value. This result quantifies the degree to which the actual operating state of the equipment deviates from the theoretical evolution benchmark, eliminates the influence of normal operating condition fluctuations, and serves as a direct data source for subsequent feature extraction and fault diagnosis.

[0058] The aforementioned time-domain evolution characteristics can refer to the dynamic change pattern of the deviation comparison results on the time axis, specifically manifested as the rate of change of the residual sequence (such as the magnitude of the first derivative) and trend pattern (such as whether it exhibits monotonic divergence); it is a key fingerprint for distinguishing the physical attributes of faults, and is used to decouple and identify the cumulative effect of slow variables and the sudden disturbance of fast variables in the time domain.

[0059] The above-mentioned fault types can refer to the abnormal attribute classification determined by the system based on the time-domain evolution characteristics. They mainly cover the gradual performance degradation (such as slow efficiency decline) caused by the long-term cumulative effect of air components (particulate matter / water), as well as the sudden mechanical failure (such as component damage) originating from the equipment itself, which directly determines whether to adopt optimization control or protection control in the future.

[0060] In some embodiments, acquiring multi-dimensional monitoring data of the compressed air energy storage power station for the current time period may specifically include:

[0061] By installing air temperature and pressure sensors and gas flow meters at the inlet section of the compressor unit, the initial temperature, pressure, and mass flow rate of the intake air are acquired in real time. Temperature and pressure sensors installed at the inlet and outlet of the compression heat exchanger and expansion heat exchanger are used to monitor the changes in the state of the medium during the compression heat release and expansion heat absorption processes. In particular, the salt cavern inlet air temperature and pressure sensor installed in the salt cavern inlet section pipeline is used to monitor the wellhead pressure and temperature fluctuations during the gas injection or gas extraction process at high frequency. The analog signals collected by all the above sensors are converted into digital signals to construct a thermodynamic state data stream characterizing the work done by the system and the fluid flow characteristics.

[0062] Simultaneously, air composition detectors installed at the compressor unit inlet and dedicated composition detectors installed at the salt cavern inlet are used to monitor the flow of microscopic materials. On one hand, detection modules based on laser scattering or beta-ray absorption principles are used to quantify the concentration of micron-sized particulate matter (such as PM2.5, PM10, and salt spray aerosols) in the air in real time to assess their cumulative impact on heat exchanger fouling and flow channel resistance. On the other hand, integrated high-precision dew point meters or humidity sensors are used to monitor the relative humidity or absolute moisture content of the air flowing through the pipeline in real time, with particular emphasis on monitoring the moisture saturation of the gas extracted from the underground salt cavern, thereby obtaining air composition data streams containing particulate matter content and moisture content data.

[0063] Finally, the heterogeneous data is spatiotemporally fused and preprocessed to generate standardized multidimensional monitoring data. The thermodynamic state data and air composition data are aggregated to an edge computing gateway or central server via a field industrial bus or Ethernet. In this stage, all data packets uploaded by sensors are timestamped using the Precise Time Protocol (PTP), and strict clock synchronization is performed to ensure that data from different dimensions are strictly aligned on the timeline. Subsequently, the raw data is cleaned to remove outliers caused by electromagnetic interference or momentary sensor malfunctions. The synchronized and cleaned thermodynamic state vector and air composition vector are then concatenated to construct a multidimensional real-time monitoring dataset containing complete system state information for the current time period. This provides high-quality data input for subsequent parameter calibration and evolutionary deduction of online simulation models.

[0064] In some embodiments, the method may further include the following:

[0065] An application is made to a monitoring system for compressed air energy storage power stations. The monitoring system includes at least: a salt cavern, a compressor unit, an expander unit, a low-temperature thermal storage tank, a high-temperature thermal storage tank, a compression-end heat exchanger, an expansion-end heat exchanger, a gas-liquid separator, and a silencer. The salt cavern is connected to the gas-liquid separator, the compressor unit, the compression-end heat exchanger, the expander unit, and the expansion-end heat exchanger via pipelines. The silencer is connected to the expander unit via pipelines. The low-temperature thermal storage tank is connected to the compression-end heat exchanger via pipelines. The high-temperature thermal storage tank is connected to the expansion-end heat exchanger via pipelines. Specifically, the method includes:

[0066] During off-peak electricity hours, outside air is controlled to enter the compressor unit. Before entering the compressor unit, the air is filtered and its composition is detected. The air is then compressed within the compressor unit and cooled by the compression-end heat exchanger before entering the gas-liquid separator for gas-liquid separation. The separated air is then injected into the salt cavern. Simultaneously, the medium in the low-temperature heat storage tank is controlled to flow through the compression-end heat exchanger to absorb heat before flowing into the high-temperature heat storage tank. During this process, the inlet air composition data of the compressor unit, the compressor unit operating status parameters, the thermodynamic parameters of the compression-end heat exchanger, and the liquid level thermodynamic parameters of the high-temperature heat storage tank are collected, and the collected data are used to construct the first monitoring dataset.

[0067] During peak electricity consumption periods, the air in the salt cavern is released under control. Before entering the expander unit, the released air is filtered and its composition is detected. The air is then heated by the expansion end heat exchanger before entering the expander unit to expand and perform work. The air after performing work is discharged through the silencer. Simultaneously, the medium in the high-temperature heat storage tank is controlled to flow through the expansion end heat exchanger to release heat before flowing into the low-temperature heat storage tank. During this process, the inlet air composition data of the expander unit, the operating status parameters of the expander unit, the thermodynamic parameters of the expansion end heat exchanger, and the liquid level thermodynamic parameters of the low-temperature heat storage tank are collected, and the collected data are used to construct a second monitoring dataset.

[0068] Based on the first monitoring dataset and the second monitoring dataset, determine the multidimensional monitoring data of the compressed air energy storage power station for the current time period.

[0069] In some embodiments, the compressed air energy storage power station monitoring system further includes: a compressor end inlet air filter and a compressor unit inlet air component detector; both the compressor end inlet air filter and the compressor unit inlet air component detector are arranged on the inlet section pipeline of the compressor unit.

[0070] In some embodiments, the compressed air energy storage power station monitoring system further includes: a compressor unit speed vibration sensor and a compressor end heat exchanger inlet and outlet temperature and pressure sensor; the compressor unit speed vibration sensor is arranged on both sides of the compressor unit shaft, and the compressor end heat exchanger inlet and outlet temperature and pressure sensor is arranged on the inlet and outlet pipelines of the compressor end heat exchanger.

[0071] In some embodiments, the compressed air energy storage power station monitoring system further includes: a high-temperature thermal storage tank temperature and pressure level sensor and a high-temperature thermal storage tank temperature and pressure sensor; the high-temperature thermal storage tank level sensor and the high-temperature thermal storage tank temperature and pressure sensor are both arranged on the tank body of the high-temperature thermal storage tank.

[0072] In some embodiments, determining the theoretical reference value for the current time period by utilizing a preset online simulation model and based on multidimensional monitoring data of the reference time period may specifically include:

[0073] A pre-defined thermodynamic digital twin evolution model is used, and multi-dimensional monitoring data within a reference time period is loaded into the model. This reference time period is typically set as a complete operating cycle immediately preceding the current moment, containing at least one complete compression period (energy storage process) and one complete expansion period (energy release process). Using this historical thermodynamic state data (such as pressure and temperature curves from the previous cycle), the model recreates the historical baseline state trajectories of key equipment in the power plant within the reference time period in virtual space, thereby establishing the initial reference frame for the system performance evolution.

[0074] Next, multiphysics state evolution calculations based on air composition data are performed. The focus is on extracting air composition data within a reference time period and performing integration calculations using the multiphysics state evolution operators integrated within the model. Specifically, based on the water content data within the reference time period, the enthalpy shift of the specific heat capacity and compressibility factor of the air working fluid after undergoing the previous compression and expansion cycle is calculated. Simultaneously, based on the particulate matter content data within the reference time period and combined with fluid dynamics principles, the cumulative increase in fouling thermal resistance at the heat exchanger interface due to impurity deposition in the previous cycle is calculated. This process aims to quantify the cumulative physical decay effect of the "material flow" (water / particles) on the "energy flow" over time.

[0075] Finally, based on the results of the above evolutionary calculations, a time step recursion and parameter correction are performed for the current moment to generate the final theoretical reference value. The calculated enthalpy drift integral and the cumulative increment of fouling thermal resistance are used as correction factors and superimposed on the aforementioned historical baseline state trajectory. The model state is then extrapolated to the current time period according to the time step. In this way, the theoretical reference value output by the model (such as the theoretically expected heat transfer efficiency or work power at the current moment) is no longer a static value based on an ideal design, but a dynamic baseline that inherits historical operating inertia and includes reasonable cumulative component effects.

[0076] In some embodiments, the method of determining a theoretical reference value for the current time period by utilizing a preset online simulation model and based on multidimensional monitoring data of the reference time period includes, in specific implementation, the method comprising:

[0077] S1: Based on the multidimensional monitoring data of the reference time period, the boundary conditions of the preset online simulation model are set using the thermodynamic state data of the reference time period to obtain the initialized model;

[0078] S2: Input the air composition data for the reference time period into the initialized model;

[0079] S3: Update the fouling thermal resistance parameters in the initialized model using particulate matter content data; update the thermophysical parameters in the initialized model using moisture content data to obtain the updated model;

[0080] S4: Based on the updated model, solve the thermodynamic equations to calculate the theoretical reference value for the current time period.

[0081] Specifically, the initialization of the simulation model's boundary conditions is involved. Multidimensional monitoring data from a reference time period (e.g., a complete charge-discharge cycle immediately preceding the current moment) is retrieved from a historical database. Thermodynamic state data, including the pressure, temperature, and flow sequences at the compressor or expander inlet, are extracted and mapped to the dynamic inlet boundary conditions of the pre-defined online simulation model. Through this process, the simulation model's operating conditions are "backtracked" and locked to the physical environment of the reference time period, resulting in an initialized model aligned with historical actual operating conditions.

[0082] Next, a deep update of the model parameters based on air composition data is performed to eliminate calculation errors caused by differences in working fluid composition. Air composition data within the reference time period is input into the initialized model, and dual-channel parameter correction is performed: On one hand, using water content data (such as relative humidity integral values) within the reference time period, combined with the thermodynamic property equation of moist air, the specific heat capacity correction factor and density correction factor of the air working fluid are calculated. Based on this, the thermophysical parameters (i.e., fluid specific heat capacity and density) in the fluid constitutive equation within the model are numerically updated to reflect the true enthalpy change characteristics of moist air within the reference time period. On the other hand, using particulate matter content data within the reference time period, the particle deposition kinetics sub-model is driven to calculate the deposition rate of fouling on the heat exchange surface. The cumulative thermal resistance value at the end of the reference time period is calculated through time integration, and the fouling thermal resistance parameters (i.e., the attenuated heat transfer coefficient) of the heat exchanger in the model are updated accordingly.

[0083] Through the above two corrections, an updated model was obtained that includes both working fluid property calibration and equipment contamination calibration. Finally, based on this updated model, the thermodynamic equations were solved and derived to generate a benchmark for the current time period. Inheriting the thermodynamic parameters and fouling thermal resistance parameters evolved from the reference time period, the thermodynamic governing equations (including mass conservation, momentum conservation, and energy conservation equations) were invoked to recursively solve the model over time. The solver calculated the theoretical thermodynamic response (such as theoretical outlet temperature and theoretical work efficiency) that should be exhibited under the current operating conditions. This calculation result is output as a theoretical reference value for the current time period, representing the ideal performance benchmark that the equipment should have under the current equipment aging level (determined by the reference time period) and the influence of the current air composition when it is in a "non-sudden failure" state.

[0084] By initializing the boundary conditions of the simulation model using thermodynamic state data from a reference time period, and by deeply updating the fouling thermal resistance and thermophysical parameters in the model based on air composition data (particulate matter and water content), theoretical reference values ​​that inherit historical evolution inertia and conform to the current working fluid characteristics can be accurately calculated.

[0085] In some embodiments, the method involves comparing the deviations based on the theoretical reference values ​​and multidimensional monitoring data for the current time period to obtain the deviation comparison results, and extracting the temporal evolution characteristics for the current time period based on the deviation comparison results. In specific implementations, the method may further include the following:

[0086] S1: Based on the theoretical reference value and the multidimensional monitoring data of the current time period, compare the deviations and determine the corresponding residual sequence;

[0087] S2: Extract time-domain features from the residual sequence to obtain the time-domain evolution features of the current time period; wherein, the time-domain evolution features include at least the rate of change feature value and the trend pattern feature.

[0088] Specifically, the measured multidimensional monitoring data collected during the current time period (e.g., measured values ​​of compressor outlet temperature and heat exchanger heat transfer coefficient) are first aligned with the theoretical reference values ​​determined in the preceding steps along the time axis. At each sampling moment, a difference operation is performed to calculate the absolute or relative deviation between the measured and theoretical values, generating a real-time residual sequence that changes continuously over time. This sequence intuitively reflects the dynamic trajectory of the equipment's current operating state deviating from the "ideal evolution benchmark," eliminating background interference caused by fluctuations in operating conditions (such as load increases and decreases) and retaining only the deviation components related to the equipment's health.

[0089] Next, the rate of change feature value extraction for the residual sequence is performed. A sliding observation window of a preset length (e.g., covering data points from the past 10 or 30 minutes) is used to extract the real-time residual sequence, and the data within the window undergoes first-order differentiation or linear regression slope calculation. This operation yields the rate of change feature value characterizing the rate of deviation development. This feature value has a clear physical orientation: a very large rate of change feature value (i.e., a steep slope) typically corresponds to the instantaneous breakage of mechanical components or sensor malfunctions; a small but constant rate of change feature value (i.e., a gentle slope) corresponds to cumulative processes such as dust accumulation or scaling.

[0090] Finally, trend pattern feature extraction is performed on the residual sequence to identify the evolution direction and stability of the deviation. The system performs waveform analysis on the residual sequence within the sliding window to identify whether it exhibits monotonicity or oscillation. Specifically, it detects whether the sequence maintains a continuous monotonically divergent trend on the time axis (i.e., the deviation value always accumulates in the positive or negative direction without any reverse pullback), which serves as a key morphological basis for judging "soft faults." At the same time, it detects whether the sequence has disordered high-frequency fluctuations around zero, thereby identifying and filtering non-fault-related random noise or transient disturbances. These two forms (monotonically divergent and disordered oscillation) together constitute the trend pattern features, providing rich criteria for subsequent accurate fault attribution.

[0091] In some embodiments, the method involves comparing the deviations based on the theoretical reference values ​​and multidimensional monitoring data for the current time period to obtain the deviation comparison results, and extracting the temporal evolution characteristics for the current time period based on the deviation comparison results. In specific implementations, the method may further include the following:

[0092] A trend test method based on time series statistics is employed. This method first involves normalizing the residual data, calculating the difference between the measured value and the theoretical reference value within the current time period, and then dividing this difference by the theoretical reference value to generate a dimensionless normalized relative residual sequence. This normalization process eliminates the influence of differences in the magnitude of basic values ​​under different power plant operating conditions (e.g., high load and low load states) on the absolute value of the deviation, ensuring that the extracted features have a uniform measurement standard across the entire operating range, facilitating subsequent statistical analysis.

[0093] Next, a preset time window is selected, and the Mann-Kendall trend test is performed on the normalized relative residual series within the window to extract trend morphology features. This is a non-parametric statistical test method that does not require the data to follow a normal distribution. The significance of the trend is determined by calculating the Z-value and the confidence level P-value of the series. Specifically, if the absolute value of the calculated Z-value is greater than the preset confidence level threshold (e.g., 1.96, corresponding to 95% confidence level) and the P-value is less than 0.05, the system determines that the series has a significant monotonic trend on the time axis, i.e., it is identified as a "monotonically divergent" trend morphology feature; conversely, if the Z-value is small or the P-value is large, it indicates that the series has no significant trend and belongs to random fluctuations around the mean, i.e., it is identified as an "oscillating convergent" trend morphology feature.

[0094] Finally, for the residual sequence within the same time window, the least squares method is used for linear regression fitting to extract the rate of change feature value. A linear regression equation for the deviation changing over time is established, and the absolute value of the slope k of this regression equation is extracted and defined as the rate of change feature value. Compared with simple two-point difference calculation, the rate of change feature based on the regression slope can effectively smooth out the interference of single-point noise and more accurately reflect the average drift rate of the deviation over a period of time. The system ultimately outputs the identified trend pattern features (monotonic or oscillating) and rate of change feature value (slope magnitude) as time-domain evolution features for subsequent multi-level fault attribution classification.

[0095] In some embodiments, the method for determining the fault type based on the temporal evolution characteristics of the current time period may further include the following:

[0096] S1: When the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a monotonically divergent state, the fault type is determined to be a gradual performance degradation fault caused by the cumulative effect of air components.

[0097] S2: When the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a non-monotonic divergent state, the fault type is determined to be a transient disturbance fault and filtered out.

[0098] S3: When the characteristic value of the rate of change is not lower than the preset mutation threshold, the fault type is determined to be a sudden mechanical fault originating from the equipment itself.

[0099] Specifically, the extracted rate of change feature values ​​are first assessed using a threshold. When the rate of change feature value is found to be below a preset abrupt change threshold, it indicates that the current deviation growth rate is relatively gradual, ruling out the possibility of severe structural damage. Based on this, a secondary verification is performed using trend morphology features: if the trend morphology features exhibit a continuous monotonically divergent state (i.e., the deviation value continuously accumulates and increases in a single direction), it is confirmed that the deviation conforms to the gradual law of physical dirt deposition or corrosion, thus identifying the fault type as a gradual performance degradation fault caused by the cumulative effect of air components. This judgment logic effectively separates "soft faults" from background fluctuations.

[0100] Secondly, for cases where the rate of change characteristic value is also below the preset abrupt change threshold, non-fault-related disturbances also need to be excluded. If the trend pattern is detected to be non-monotonic and divergent (e.g., exhibiting random oscillations around zero, convergent fluctuations, or disordered jumps), it indicates that the deviation lacks directional physical evolution and is usually caused by sensor electromagnetic noise, transient airflow disturbances, or communication packet loss. Therefore, such cases are identified as transient disturbance faults and filtered or ignored in subsequent control decisions, i.e., no substantial parameter adjustments or shutdown actions are triggered, thereby greatly reducing the false alarm rate of the system and ensuring the robustness of the control strategy.

[0101] Finally, when the rate of change characteristic value is identified as not lower than (i.e., greater than or equal to) a preset abrupt change threshold, it indicates that the deviation has undergone a drastic step or pulse-like change within a very short period of time. This signal characteristic usually violates the physical laws of natural wear or dust accumulation, but highly matches the characteristics of structural failures of mechanical components (such as blade breakage, bearing seizure, valve stem detachment, etc.). Therefore, regardless of the trend pattern at this time, the fault type is preferentially identified as a sudden mechanical failure originating from the equipment itself. This judgment logic ensures the highest priority response to high-risk faults, providing a decision basis for immediately triggering the safety blocking mode.

[0102] In some embodiments, the method of performing corresponding operation control on the compressed air energy storage power station according to the fault type may further include the following:

[0103] When the fault type is the gradual performance degradation fault, adjust the first speed setting value of the low-temperature circulating pump installed on the connecting pipeline between the low-temperature heat storage tank and the compression end heat exchanger, and / or adjust the second speed setting value of the high-temperature circulating pump installed on the connecting pipeline between the high-temperature heat storage tank and the expansion end heat exchanger, and monitor the heat exchange efficiency of the high-temperature circulating pump and / or the low-temperature circulating pump until the heat exchange efficiency reaches the preset deviation range.

[0104] Specifically, when the fault type is determined to be a gradual performance degradation fault caused by the cumulative effect of air components (such as ash or scale buildup on the heat exchanger surface), the intelligent monitoring system will automatically activate the preset "parameter compensation mode". In this mode, the system's control objective is no longer a simple fault alarm, but rather to actively adjust the thermal-hydraulic parameters to forcibly maintain the overall heat exchange efficiency of the system even when there is slight fouling thermal resistance in the equipment, so as to achieve efficient operation "with faults".

[0105] Next, specific speed adjustment actions are executed for the circulation loop of the heat storage medium. Based on the current efficiency deviation, the system generates speed correction commands for the circulation pumps. Specifically, if the heat exchange efficiency decreases during compression, the system controls the low-temperature circulation pump connected to the low-temperature heat storage tank and the compression end heat exchanger, increasing its first speed setting. If the heat exchange efficiency decreases during expansion, the system controls the high-temperature circulation pump connected to the high-temperature heat storage tank and the expansion end heat exchanger, increasing its second speed setting. By increasing the pump speed, the mass flow rate of the heat storage medium (such as water or heat transfer oil) flowing through the heat exchanger is increased, thereby utilizing the increased convective heat transfer coefficient resulting from the increased flow velocity to dynamically offset the increase in thermal resistance caused by fouling.

[0106] Finally, closed-loop monitoring and adjustment termination determination based on heat exchange efficiency feedback are executed. While adjusting the speed, the system continuously uses sensor data to calculate the real-time heat exchange efficiency of the loop containing the high-temperature and / or low-temperature circulating pumps, and compares it with the theoretical operating baseline value. The system monitors the recovery of this heat exchange efficiency in real time. If the efficiency deviation still exceeds the allowable range, the speed is further fine-tuned until the heat exchange efficiency recovers and stabilizes within the preset deviation range (e.g., within ±2% of the theoretical value). At this point, the system locks the current speed setpoint, completing this thermal compensation control. This closed-loop mechanism ensures the accuracy of compensation, avoiding both efficiency loss due to under-adjustment and pump power waste caused by over-adjustment.

[0107] In some embodiments, the method of performing corresponding operation control on the compressed air energy storage power station according to the fault type may further include the following:

[0108] When the fault type is the sudden mechanical fault, the first switch valve located at the salt cave inlet section and the second switch valve located at the salt cave outlet section are controlled to perform a rapid closing action, and the power input of the compressor unit or expander unit is cut off simultaneously.

[0109] Specifically, when the intelligent monitoring system determines the fault type as a sudden mechanical failure originating from the equipment itself through time-domain evolution characteristic analysis (e.g., detecting a sharp increase in the rate of change of deviation characteristic value exceeding a preset mutation threshold), the system will immediately determine that the power station is in a dangerous state. At this time, the system automatically bypasses the conventional PID control logic and prioritizes the activation of the highest priority "safety blocking mode" to prevent the further escalation of the fault consequences.

[0110] Next, a physical isolation operation is performed on the salt cavern gas storage facility. The system generates an emergency shutdown command through an interlocking protection subsystem, which is sent directly to the key valve actuators deployed on the salt cavern pipeline. Specifically, the system controls the first switching valve at the salt cavern inlet and the second switching valve at the salt cavern outlet to perform rapid closing actions. Since the salt cavern is a huge energy source storing high-pressure air, the rapid shut-off of these two valves (typically requiring completion within seconds) can quickly disconnect the gas source from the ground equipment, preventing high-pressure gas leaks or blowouts caused by ruptures in ground equipment (such as heat exchangers or pipelines), thus achieving a physical blockade of the core energy storage area.

[0111] Finally, while performing the gas path blocking, the system simultaneously executes a power cut-off operation for the rotating machinery. The system sends a trip command to the drive motor of the compressor unit or the intake regulating valve of the expander unit, simultaneously cutting off their power input. For example, for a running compressor, the system disconnects its main motor power; for a running expander, the system forcibly closes its intake quick-closing valve. This synchronous action aims to eliminate the rotational inertia and driving force of the mechanical equipment, preventing equipment disintegration or uncontrolled vibration caused by mechanical failures such as blade breakage or bearing seizure, thereby strictly limiting the impact of the fault to a local area and preventing the cascading spread of the fault to the entire system.

[0112] In some embodiments, the method may further include the following:

[0113] S1: Using the parallel time-series feature extraction module of the preset thermodynamic digital twin evolution model, determine the corresponding time-series feature vector sequence based on the multi-dimensional monitoring data of the reference time period;

[0114] S2: Using the physical constraint decoding module of the preset thermodynamic digital twin evolution model, under the constraint of thermodynamic energy conservation boundary conditions, the theoretical reference value of the current time period is determined according to the time-series feature vector sequence;

[0115] The preset thermodynamic digital twin evolution model is a deep neural network model based on a heterogeneous dual-stream architecture; the parallel temporal feature extraction module is a module based on a bidirectional long short-term memory network structure with independent dual-channel encoding; and the physical constraint decoding module is a module based on a multilayer perceptron and energy manifold correction structure.

[0116] Specifically, the system first invokes a pre-defined thermodynamic digital twin evolution model, which is constructed as a deep neural network model based on a heterogeneous dual-flow architecture, designed to simultaneously process physical data of different properties in compressed air energy storage. The model then utilizes its integrated parallel temporal feature extraction module to process multidimensional monitoring data within a reference time period (i.e., the complete historical operating cycle). This module employs a bidirectional long short-term memory (Bi-LSTM) network structure based on independent dual-channel encoding: one channel is dedicated to encoding thermodynamic state data (such as transient fluctuations in temperature and pressure), using the bidirectional recurrent units of the Bi-LSTM to capture the dynamic inertia of energy conversion; the other independent channel is dedicated to encoding air component data (such as particulate matter and moisture content), capturing the long-term evolutionary trend of impurity accumulation. Through this dual-channel parallel processing, the model maps the raw data within the reference time period into a sequence of temporal feature vectors containing high-dimensional physical semantics.

[0117] Next, the decoding and prediction operations based on physical mechanisms are performed. The generated time-series feature vector sequence is input into the physical constraint decoding module of the model. This module does not employ a typical unconstrained regression layer, but rather a composite architecture based on a multilayer perceptron (MLP) and an energy manifold correction structure. First, the MLP performs a nonlinear mapping on the input feature vectors to generate preliminary predictions; subsequently, the data stream enters the energy manifold correction layer. This layer incorporates the thermodynamic energy conservation boundary conditions of the compressed air energy storage system (e.g., the balance equation between compression work and the change in internal energy), serving as a mandatory mathematical constraint plane (manifold).

[0118] Finally, during the decoding process, the model calculates the residual between the initial predicted value and the energy conservation boundary, and forces the output result to be projected into the physically feasible region that satisfies the law of energy conservation through gradient descent or the Lagrange multiplier method. The value output after this correction process is the theoretical reference value for the current time period. This mechanism ensures that even in a purely data-driven neural network, the derived theoretical benchmark value strictly follows the first law of thermodynamics, thus providing a highly confident benchmark for subsequent bias comparisons.

[0119] As can be seen from the above, the monitoring and control method for a compressed air energy storage power station provided in this specification acquires multi-dimensional monitoring data of the compressed air energy storage power station for the current time period and multi-dimensional monitoring data for a reference time period. The multi-dimensional monitoring data includes thermodynamic state data and air composition data. The air composition data includes at least particulate matter content data and moisture content data. The reference time period includes at least one consecutive compression time period and expansion time period. Using a preset online simulation model, a theoretical reference value for the current time period is determined based on the multi-dimensional monitoring data of the reference time period. A deviation comparison is performed between the theoretical reference value and the multi-dimensional monitoring data for the current time period to obtain a deviation comparison result. Based on the deviation comparison result, the time-domain evolution characteristics of the current time period are extracted. Based on the time-domain evolution characteristics of the current time period, the fault type is determined. And based on the fault type, corresponding operation control is performed on the compressed air energy storage power station. In this way, by acquiring air composition data including particulate matter content and moisture content, and incorporating this air composition data into multidimensional monitoring data, the monitoring dimensions of the system are effectively expanded. Simultaneously, this method utilizes a pre-set online simulation model to accurately determine the theoretical reference value for the current time period based on multidimensional monitoring data containing at least one consecutive compression and expansion time period. Furthermore, it extracts time-domain evolution features based on deviation comparison results to automatically determine the fault type and implement corresponding operational control. This achieves timely fault identification and precise fault management, significantly solving the technical shortcomings of delayed control response time and high overall operation and maintenance costs.

[0120] See Figure 2 As shown in the embodiments of this specification, a specific electronic device is also provided, wherein the electronic device includes a network communication port 201, a processor 202 and a memory 203, and the above structures are connected by internal cables so that the various structures can perform specific data interaction.

[0121] Specifically, the network communication port 201 can be used to acquire multi-dimensional monitoring data of the compressed air energy storage power station for the current time period, as well as multi-dimensional monitoring data for a reference time period. The multi-dimensional monitoring data includes thermodynamic state data and air composition data. The air composition data includes at least particulate matter content data and moisture content data. The reference time period includes at least one consecutive compression time period and expansion time period.

[0122] Specifically, the processor 202 can be used to determine the theoretical reference value for the current time period based on the multi-dimensional monitoring data of the reference time period using a preset online simulation model; compare the deviation between the theoretical reference value and the multi-dimensional monitoring data of the current time period to obtain the deviation comparison result; extract the time-domain evolution characteristics of the current time period based on the deviation comparison result; determine the fault type based on the time-domain evolution characteristics of the current time period; and perform corresponding operation control on the compressed air energy storage power station based on the fault type.

[0123] The memory 203 can be used to store the corresponding instruction program.

[0124] Based on the above method, the relevant structural performance of electronic equipment can be effectively utilized to improve the data processing speed of electronic equipment and efficiently realize a monitoring and control method based on compressed air energy storage power station.

[0125] In this embodiment, the network communication port 201 can be a virtual port bound to different communication protocols, thereby enabling the sending or receiving of different data. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. Furthermore, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM or CDMA; it can also be a Wi-Fi chip; or it can be a Bluetooth chip.

[0126] In this embodiment, the processor 202 can be implemented in any suitable manner. For example, the processor can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers, etc. This specification is not limiting.

[0127] In this embodiment, the memory 203 may include a hierarchy. In a digital system, anything that can store binary data can be a memory. In an integrated circuit, a circuit with storage function but no physical form is also called a memory, such as RAM, FIFO, etc. In a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0128] This specification also provides a computer-readable storage medium based on the above-described monitoring and control method for a compressed air energy storage power station. This medium acquires multi-dimensional monitoring data of the compressed air energy storage power station for the current time period and multi-dimensional monitoring data for a reference time period. The multi-dimensional monitoring data includes thermodynamic state data and air composition data. The air composition data includes at least particulate matter content data and moisture content data. The reference time period includes at least one consecutive compression time period and expansion time period. Using a preset online simulation model, a theoretical reference value for the current time period is determined based on the multi-dimensional monitoring data of the reference time period. A deviation comparison is performed between the theoretical reference value and the multi-dimensional monitoring data for the current time period to obtain a deviation comparison result. Based on the deviation comparison result, the time-domain evolution characteristics of the current time period are extracted. Based on the time-domain evolution characteristics of the current time period, the fault type is determined. And based on the fault type, corresponding operational control is performed on the compressed air energy storage power station.

[0129] In this embodiment, the storage medium includes, but is not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions. The network communication unit can be an interface configured according to standards specified in the communication protocol for network connection communication.

[0130] In this embodiment, the specific functions and effects implemented by the program instructions stored in the computer-readable storage medium can be explained in comparison with other embodiments, and will not be repeated here.

[0131] See Figure 3 At the software level, this specification also provides a monitoring and control device based on a compressed air energy storage power station, which may specifically include the following structural modules:

[0132] The data acquisition module 301 is used to acquire multi-dimensional monitoring data of the compressed air energy storage power station for the current time period and multi-dimensional monitoring data for a reference time period; wherein, the multi-dimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period;

[0133] The reference value determination module 302 is used to determine the theoretical reference value for the current time period by using a preset online simulation model and based on the multidimensional monitoring data of the reference time period.

[0134] The feature extraction module 303 is used to compare the deviations based on the theoretical reference values ​​and multidimensional monitoring data of the current time period, obtain the deviation comparison results, and extract the temporal evolution features of the current time period based on the deviation comparison results.

[0135] The operation control module 304 is used to determine the fault type based on the time-domain evolution characteristics of the current time period; and to perform corresponding operation control on the compressed air energy storage power station according to the fault type.

[0136] In some embodiments, the reference value determination module 302, in specific implementation, sets the boundary conditions of a preset online simulation model using the thermodynamic state data of the reference time period based on the multidimensional monitoring data of the reference time period to obtain an initialized model; inputs the air composition data of the reference time period into the initialized model; updates the fouling thermal resistance parameters in the initialized model using particulate matter content data; updates the thermophysical parameters in the initialized model using moisture content data to obtain an updated model; and solves the thermodynamic equations based on the updated model to calculate the theoretical reference value for the current time period.

[0137] In some embodiments, the feature extraction module 303, in its specific implementation, performs a deviation comparison based on the theoretical reference value of the current time period and the multidimensional monitoring data of the current time period to determine the corresponding residual sequence; performs time-domain feature extraction on the residual sequence to obtain the time-domain evolution features of the current time period; wherein, the time-domain evolution features include at least the rate of change feature value and the trend pattern feature.

[0138] In some embodiments, the fault type determination execution step in the above-mentioned operation control module 304 is specifically implemented as follows: when the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a monotonically divergent state, the fault type is determined to be a gradual performance degradation fault caused by the cumulative effect of air components; when the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a non-monotonic divergent state, the fault type is determined to be a transient disturbance fault and filtered out; when the change rate characteristic value is not lower than the preset mutation threshold, the fault type is determined to be a sudden mechanical fault originating from the equipment body.

[0139] In some embodiments, the above-mentioned operation control module 304, when the fault type is the gradual performance degradation fault, adjusts the first speed setting value of the low-temperature circulating pump installed on the connecting pipeline between the low-temperature heat storage tank and the compression end heat exchanger, and / or adjusts the second speed setting value of the high-temperature circulating pump installed on the connecting pipeline between the high-temperature heat storage tank and the expansion end heat exchanger, and monitors the heat exchange efficiency of the high-temperature circulating pump and / or the low-temperature circulating pump until the heat exchange efficiency reaches a preset deviation range. When the fault type is the sudden mechanical fault, it controls the first switching valve installed in the salt cavern inlet section and the second switching valve installed in the salt cavern outlet section to perform a rapid closing action, and simultaneously cuts off the power input of the compressor unit or the expander unit.

[0140] In some embodiments, the above-described device further includes: a parallel temporal feature extraction module using a preset thermodynamic digital twin evolution model to determine a corresponding temporal feature vector sequence based on multidimensional monitoring data of the reference time period; and a physical constraint decoding module using a preset thermodynamic digital twin evolution model to determine a theoretical reference value for the current time period based on the temporal feature vector sequence under the constraint of thermodynamic energy conservation boundary conditions; wherein the preset thermodynamic digital twin evolution model is a deep neural network model based on a heterogeneous dual-stream architecture; the parallel temporal feature extraction module is a module based on a bidirectional long short-term memory network structure with independent dual-channel encoding; and the physical constraint decoding module is a module based on a multilayer perceptron and energy manifold correction structure.

[0141] It should be noted that the units, devices, or modules described in the above embodiments can be implemented by computer chips or physical entities, or by products with certain functions. For ease of description, the above devices are described by dividing them into various modules according to their functions. Of course, in implementing this specification, the functions of each module can be implemented in the same software and / or hardware, or modules that implement the same function can be implemented by a combination of sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, units or components can be combined or integrated into another system, or some features can be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection between the devices or units shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0142] As can be seen from the above, the monitoring and control device for a compressed air energy storage power station provided in the embodiments of this specification acquires multi-dimensional monitoring data of the compressed air energy storage power station for the current time period and multi-dimensional monitoring data for a reference time period. The multi-dimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period. By utilizing a preset online simulation model, a theoretical reference value for the current time period is determined based on the multi-dimensional monitoring data of the reference time period. A deviation comparison is performed between the theoretical reference value and the multi-dimensional monitoring data for the current time period to obtain a deviation comparison result. Based on the deviation comparison result, the time-domain evolution characteristics of the current time period are extracted. Based on the time-domain evolution characteristics of the current time period, the fault type is determined. And based on the fault type, corresponding operation control is performed on the compressed air energy storage power station.

[0143] In a specific scenario example, the monitoring and control method and device for a compressed air energy storage power station provided in this specification can be applied, solving the problems of limited monitoring range, reliance on manual control leading to delayed response time, and high overall operation and maintenance costs in existing compressed air energy storage power station monitoring systems. The specific implementation process may include the following:

[0144] In some embodiments, see Figure 4 and Figure 5 As shown, a monitoring system based on a compressed air energy storage power station includes a salt cavern 1, a compressor unit 2, an expander unit 3, a low-temperature heat storage tank 4, a high-temperature heat storage tank 5, a compression end heat exchanger 6, an expansion end heat exchanger 7, a gas-liquid separator 8, and a silencer 9. The salt cavern 1 is connected to the gas-liquid separator 8, the compressor unit 2, the compression end heat exchanger 6, the expander unit 3, and the expansion end heat exchanger 7 via pipelines. The silencer 9 is connected to the expander unit 3 via pipelines. The low-temperature heat storage tank 4 and the high-temperature heat storage tank 5 are connected to the compression end heat exchanger 6 and the expansion end heat exchanger 7 via pipelines.

[0145] The salt cavern 1 and its inlet and outlet sections include a first switching valve 11, a second switching valve 12, a salt cavern inlet air temperature and pressure sensor 13, and a salt cavern inlet air composition detector 14; the first switching valve 11 and the second switching valve 12 are respectively installed at the inlet and outlet sections of the salt cavern 1; the salt cavern inlet air temperature and pressure sensor 13 and the salt cavern inlet air composition detector 14 are installed at the inlet section of the salt cavern 1.

[0146] The compressor unit 2 and its inlet section include a compressor-end inlet air filter 101, a compressor-end gas flow meter 102, a compressor-end inlet air component detector 103, a compressor-end inlet air temperature and pressure sensor 21, and a compressor-end speed and vibration sensor 22. The compressor-end inlet air filter 101 and the compressor-end gas flow meter 102 are installed in the inlet section of the compressor unit 2 and connected by a connecting pipe. The compressor-end inlet air component detector 103 and the compressor-end inlet air temperature and pressure sensor 21 are installed in the inlet section pipe of the compressor unit 2. The compressor-end speed and vibration sensor 22 is installed on both sides of the compressor unit 2 body and shaft.

[0147] The expander unit 3 and its outlet section include an expander unit speed vibration sensor 31 and a silencer 9; the expander unit speed vibration sensor 31 is arranged on both sides of the expander unit 3 body and shaft; the silencer 9 is arranged in the outlet section of the expander unit 3 and is connected by a connecting pipe.

[0148] The piping loop of the cryogenic thermal storage tank 4 includes a cryogenic circulating pump 41, a first regulating valve 42, a first flow meter 43, a cryogenic temperature and pressure sensor 44, and a cryogenic level gauge 45; the cryogenic circulating pump 41 and the first regulating valve 42 are installed on the connecting pipeline between the cryogenic thermal storage tank 4 and the compression end heat exchanger 6; the first flow meter 43 is installed on the connecting pipeline between the compression end heat exchanger 6 and the high-temperature thermal storage tank 5; the cryogenic temperature and pressure sensor 44 and the cryogenic level gauge 45 are installed on the surface or inside the cryogenic thermal storage tank 4.

[0149] The piping circuit of the high-temperature thermal storage tank 5 includes a high-temperature circulating pump 51, a second regulating valve 52, a second flow meter 53, a high-temperature temperature and pressure sensor 54, and a high-temperature liquid level gauge 55; the high-temperature circulating pump 51 and the second regulating valve 52 are arranged on the connecting pipeline between the high-temperature thermal storage tank 5 and the expansion end heat exchanger 7; the second flow meter 53 is arranged on the connecting pipeline between the expansion end heat exchanger 7 and the low-temperature thermal storage tank 4; the high-temperature temperature and pressure sensor 54 and the high-temperature liquid level gauge 55 are arranged on the surface or inside the high-temperature thermal storage tank 5.

[0150] The compression end heat exchanger 6 and its inlet and outlet sections include compression end heat exchanger inlet and outlet temperature and pressure sensors 61; the compression end heat exchanger inlet and outlet temperature and pressure sensors 61 are arranged on the connecting pipeline between the compression end heat exchanger 6 and the compressor unit 2 and the steam-water separator 8.

[0151] The expansion end heat exchanger 7 and its inlet and outlet sections include an expansion end inlet air filter 201, an expansion end gas flow meter 202, an expansion unit inlet air component detector 203, and expansion end heat exchanger inlet and outlet temperature and pressure sensors 71. The expansion end inlet air filter 201 and the expansion end gas flow meter 202 are installed at the inlet section of the expansion end heat exchanger 7 and connected by a connecting pipe. The expansion unit inlet air component detector 203 is installed in the inlet section pipe of the expansion end heat exchanger 7. The expansion end heat exchanger inlet and outlet temperature and pressure sensors 71 are installed on the connecting pipe between the expansion end heat exchanger 7 and the expansion end gas flow meter 202.

[0152] The gas-liquid separator 8 includes a gas-liquid separator level gauge 81; the gas-liquid separator level gauge 81 is installed on the surface or inside the tank of the gas-liquid separator 8.

[0153] In some embodiments, during periods of low electricity demand, the compressed air energy storage power station absorbs excess electricity from the grid and draws outside air into the system. The air first passes through an inlet air filter at the compressor end to reduce particulate matter and moisture content. After passing through a gas flow meter, a compressor inlet air component detector 103, and a temperature and pressure sensor, it enters the compressor unit 2 for staged compression, accompanied by cooling from the compressor end heat exchanger 6. This completes the energy conversion from electrical energy to the internal energy of the air and the thermal energy of the high-temperature heat storage medium. During this process, the temperature and pressure sensors at the inlet and outlet of the heat exchanger simultaneously monitor the air state parameters, and the compressor unit speed and vibration sensor 22 simultaneously monitors the compressor unit. 2. During operation, the air compressed to a specified pressure continues to be further dried by the gas-liquid separator 8 and stored in the salt cavern 1 by the first switch valve 11. During this period, the air temperature and pressure sensor 13 and the component detector at the salt cavern inlet are installed to monitor the state and composition of the injected gas in real time. At the same time, the high-temperature heat storage medium that absorbs the heat of gas compression enters the high-temperature heat storage tank 5 through the first flow meter 43. The latter is equipped with a high-temperature temperature and pressure sensor 54 and a high-temperature liquid level gauge 55 to monitor the state and liquid level of the high-temperature heat storage medium in real time. The power plant intelligent monitoring system summarizes the entire process of air intake to compression into the salt cavern 1 online.

[0154] During peak electricity demand, the compressed air energy storage power station releases high-pressure air from the salt cavern 1. This high-pressure gas first passes through the second switch valve 12 and enters the inlet air filter at the expansion end to remove particulate matter and moisture carried out from the salt cavern 1. After passing through the gas flow meter and the air composition detector 203 at the inlet of the expander unit, it expands in stages through the expander unit 3, accompanied by the heat exchanger 7 at the expansion end preheating the air at the expander inlet. This allows the air to absorb the high-temperature heat storage medium and expand to do work, thereby driving the generator to generate electricity. During this process, the temperature and pressure sensors at the inlet and outlet of the heat exchanger simultaneously monitor the air state parameters, and the speed and vibration sensor 31 of the expander unit simultaneously monitors the operating status of the compressor unit 2. The air after doing work is discharged into the atmosphere through the silencer 9. The high-temperature heat storage medium cools down after releasing heat and becomes a low-temperature heat storage medium. It enters the low-temperature heat storage tank 4 through the second flow meter 53. The latter is equipped with a low-temperature temperature and pressure sensor 44 and a low-temperature liquid level gauge 45 to monitor the status and liquid level of the low-temperature heat storage medium in real time. The power station's intelligent monitoring system also summarizes the entire process of air discharge from the salt cavern 1 to the atmosphere online.

[0155] In some embodiments, see Figure 6 As shown, an intelligent control method for a monitoring system based on a compressed air energy storage power station includes closed-loop monitoring and control S1 and manual control S2.

[0156] The closed-loop monitoring and control S1 includes: S11, during power plant start-up and shutdown, various sensor devices deployed in the compressed air energy storage system upload data information such as equipment status and thermodynamic state of the thermal / energy storage medium to the data center via wired / wireless communication; S121, simultaneously establishing an online simulation model containing all key equipment of the compressed air energy storage system; S12, based on the online simulation model and the real-time status data stream of the compressed air energy storage system, calculating the theoretical operating parameters of each key device and uploading them to the data center; S13, collecting, organizing, storing, and cleaning various operation and maintenance data streams from sensor devices and the online simulation system, and simultaneously generating historical operation and maintenance data. According to the database; S14, establish an intelligent operation and maintenance platform integrating functions such as intelligent operation, intelligent monitoring, and intelligent equipment. Combine big data analysis, artificial intelligence, expert knowledge base, digital twin and other technologies to conduct in-depth comparative analysis of various operation and maintenance data streams from sensors and online simulation systems, and simultaneously generate multi-dimensional monitoring and predictive early warning reports; S15, the intelligent operation and maintenance platform, through various operation auxiliary systems such as APS unit self-start and stop system, interlock protection system, AGC and primary frequency regulation optimization control system, charging and discharging gas volume control system and thermal storage medium temperature optimization system, based on the intelligent analysis of system operation data streams and reports, combined with external adjustments and arrangements, performs closed-loop / open-loop control of each subsystem.

[0157] The manual control S2 includes: the manual operation and maintenance side, which, through comprehensive macro-level demand planning and real-time system reports, makes necessary operational intervention adjustments and equipment maintenance plans for the pressure storage system to ensure safe and efficient system operation.

[0158] In some embodiments, in the closed-loop monitoring mode of the compression process of the monitoring system of the compressed air energy storage power station, the first switching valve 11, the first regulating valve 42, and the cryogenic circulating pump 41 are opened, while other valves and the circulating pump are closed, enabling the compressed air pipeline at the compressed end and the heat storage medium pipeline between the cryogenic heat storage tank and the high-temperature heat storage tank of the compressed air energy storage system to operate; the compressed end inlet air filter 101 is used to reduce the particulate matter content and water content of the inlet air, the compressed end gas flow meter 102 is used to monitor the inlet air flow rate at the compressed end, the first flow meter 43 is used to monitor the high-temperature heat storage medium flow rate at the outlet of the compressed end heat exchanger, the compressor unit inlet air component detector 103 is used to monitor the particulate matter content and water content of the compressor unit inlet air, and the salt cavern inlet air component detector 14 is used to monitor the salt cavern injection... The temperature and pressure sensors 21 and 61 at the inlet and outlet of the compressor heat exchanger are used to monitor the temperature and pressure of the air at the compressor inlet and the air at the inlet and outlet of the compressor heat exchanger, respectively. The temperature and pressure sensor 13 at the inlet of the salt cavern is used to monitor the temperature and pressure of the gas injected into the salt cavern. The temperature and pressure sensors 44 and 54 at low temperature and high temperature are used to monitor the temperature and pressure of the heat storage medium in the low temperature heat storage tank and the high temperature heat storage tank, respectively. The speed and vibration sensor 22 of the compressor unit is used to monitor the speed and vibration of the compressor unit during operation. The level gauge 81, the low temperature level gauge 45, and the high temperature level gauge 55 of the gas-liquid separator are used to monitor the water level in the gas-liquid separator and the level of the heat storage medium in the low temperature heat storage tank and the high temperature heat storage tank, respectively.

[0159] During the operation of the power plant's compressor end, the closed-loop monitoring mode of the intelligent monitoring system also operates synchronously. The process includes: S11, the various sensor devices deployed at the compressor end upload data such as equipment status and thermodynamic state of the thermal / energy storage medium to the data center via wired / wireless communication; S12, simultaneously establishing an online simulation model containing all key equipment of the compressed air energy storage system; S12, based on the online simulation model and the real-time status data stream of the compressed air energy storage system, calculating the theoretical operating parameters of each key device and uploading them to the data center; S13, collecting, organizing, storing, and cleaning various operation and maintenance data streams from the sensor devices and the online simulation system. S14. Establish a smart operation and maintenance platform integrating functions such as intelligent operation, intelligent monitoring, and smart equipment. Combining big data analysis, artificial intelligence, expert knowledge base, digital twin and other technologies, conduct in-depth comparative analysis of various operation and maintenance data streams from sensors and online simulation systems, and simultaneously generate multi-dimensional monitoring and predictive early warning reports; S15. The smart operation and maintenance platform uses various operation auxiliary systems such as APS unit self-start and stop system, interlock protection system, AGC and primary frequency regulation optimization control system, charging and discharging gas volume control system and thermal storage medium temperature optimization system to perform closed-loop control of compressed air energy storage system based on intelligent analysis of system operation data streams and reports.

[0160] In this mode, the compression end of the compressed air energy storage system and the intelligent monitoring system work together. The former's operating status is quantified by various sensing devices and received by the latter. Through the process chain of comparative analysis with online simulation models, noise reduction and storage in the data center, analysis and early warning by the intelligent platform, and intervention and response by the multi-control system, the system can monitor and control the entire process and all aspects of air pretreatment, compression and post-treatment, ensuring the stable and efficient operation of the power station.

[0161] In some embodiments, in the open-loop monitoring mode of the expansion process of the compressed air energy storage power station's monitoring system, the second switching valve 12, the second regulating valve 52, and the high-temperature circulating pump 51 are opened, while other valves and the circulating pump are closed, enabling the expansion end air pipeline and the high-temperature heat storage tank-low-temperature heat storage tank heat storage medium pipeline of the compressed air energy storage system to operate; the expansion end inlet air filter device 201 is used to reduce the particulate matter content and water content of the salt cavern gas, the expansion end gas flow meter 202 is used to monitor the salt cavern gas flow rate, and the second flow meter 53 is used to monitor the low-temperature heat storage medium flow rate at the expansion end heat exchanger outlet. The expander unit inlet air component detector 203 is used to monitor the particulate matter content and water content of the gas collected from the salt cavern; the expansion end heat exchanger inlet and outlet temperature and pressure sensor 71 is used to monitor the temperature and pressure of the air at the inlet and outlet of the expansion end heat exchanger; the high temperature and pressure sensor 54 and the low temperature and pressure sensor 44 are used to monitor the temperature and pressure of the heat storage medium in the high temperature heat storage tank and the low temperature heat storage tank, respectively; the expander unit speed and vibration sensor 22 is used to monitor the speed and vibration of the expander unit during operation; and the high temperature level gauge 55 and the low temperature level gauge 45 are used to monitor the liquid level of the heat storage medium in the high temperature heat storage tank and the low temperature heat storage tank, respectively.

[0162] During operation at the expansion end of the power plant, the open-loop monitoring mode of the intelligent monitoring system also operates synchronously. The process includes: S11, the various sensor devices deployed at the compression end upload data such as equipment status and thermodynamic state of the thermal / energy storage medium to the data center via wired / wireless communication; S12, simultaneously establishing an online simulation model containing all key equipment of the compressed air energy storage system; S12, based on the online simulation model and the real-time status data stream of the compressed air energy storage system, calculating the theoretical operating parameters of each key device and uploading them to the data center; S13, collecting, organizing, storing, and cleaning various operation and maintenance data streams from sensor devices and the online simulation system, simultaneously generating a historical operation and maintenance database; S14, establishing an intelligent operation and maintenance platform integrating intelligent operation, intelligent monitoring, and intelligent equipment functions. The platform, combining big data analytics, artificial intelligence, expert knowledge bases, and digital twin technologies, performs in-depth comparative analysis of various operation and maintenance data streams originating from sensors and online simulation systems, simultaneously generating multi-dimensional monitoring and predictive early warning reports; S2, the manual operation and maintenance side, through comprehensive macro-level demand arrangements and real-time system reports, makes necessary operational intervention adjustments and equipment maintenance plans for the compressed air energy storage system, and directly affects the control operation auxiliary system to ensure safe and efficient system operation; S15, the intelligent operation and maintenance platform, through various operation auxiliary systems such as the APS unit self-start and stop system, interlock protection system, AGC and primary frequency regulation optimization control system, charging and discharging gas volume control system, and thermal storage medium temperature optimization system, achieves open-loop control of the compressed air energy storage system based on intelligent analysis of system operation data streams and reports, combined with adjustments and arrangements on the manual operation and maintenance side;

[0163] In this mode, the expansion end of the compressed air energy storage system and the intelligent monitoring system operate in tandem. At this time, the latter mainly uses various advanced digital technologies to deeply analyze and summarize the power plant system operation data flow and theoretical model calculation results, and generate multi-dimensional monitoring and prediction early warning reports, which are sent to the manual operation and maintenance side. This enables operation and maintenance personnel to make reasonable decisions based on information such as the power plant's operating status and the macro-conditions of power grid supply and demand, and to intervene in the closed-loop control chain of the power plant's intelligent monitoring system. While meeting the diverse operational needs of the power plant, it also maintains the economy and safety of power plant operation and maintenance.

[0164] While this specification provides the steps of operation for the methods described in the embodiments or flowcharts, more or fewer steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible order of execution among many steps and does not represent the only possible order. In actual device or client product execution, the methods shown in the embodiments or drawings may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded. The terms "first," "second," etc., are used to denote names and do not indicate any particular order.

[0165] Those skilled in the art will also know that, besides implementing the controller using purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller function as logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices within it used to implement various functions can also be considered structures within that hardware component. Alternatively, the devices used to implement various functions can be considered as both software modules implementing the method and structures within a hardware component.

[0166] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this specification can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of this specification can essentially be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments of this specification.

[0167] Although this specification has been described by way of examples, those skilled in the art will recognize that many variations and modifications are possible without departing from the spirit of this specification, and it is intended that the appended claims cover such variations and modifications without departing from the spirit of this specification.

Claims

1. A monitoring and control method for a compressed air energy storage power station, characterized in that, include: Acquire multidimensional monitoring data of the compressed air energy storage power station for the current time period, as well as multidimensional monitoring data for a reference time period; wherein, the multidimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period; By utilizing a preset online simulation model and based on multidimensional monitoring data from the reference time period, a theoretical reference value for the current time period is determined. Based on the theoretical reference value and the multidimensional monitoring data of the current time period, a deviation comparison is performed to obtain the deviation comparison result. Based on the deviation comparison result, the temporal evolution characteristics of the current time period are extracted. Based on the time-domain evolution characteristics of the current time period, the fault type is determined; and the compressed air energy storage power station is operated accordingly based on the fault type.

2. The method according to claim 1, characterized in that, The step of determining theoretical reference values ​​for the current time period by utilizing a preset online simulation model and based on multidimensional monitoring data from the reference time period includes: Based on the multidimensional monitoring data of the reference time period, the boundary conditions of the preset online simulation model are set using the thermodynamic state data of the reference time period to obtain the initialized model; Input the air composition data for the reference time period into the initialized model; The fouling thermal resistance parameters in the initial model were updated using particulate matter content data; the thermophysical parameters in the initial model were updated using moisture content data, resulting in the updated model. Based on the updated model, the thermodynamic equations are solved to calculate the theoretical reference values ​​for the current time period.

3. The method according to claim 2, characterized in that, The process involves comparing the theoretical reference value and the multidimensional monitoring data for the current time period to obtain the comparison result, and then extracting the temporal evolution characteristics of the current time period based on the comparison result, including: Based on the theoretical reference value for the current time period and the multidimensional monitoring data for the current time period, the deviation is compared to determine the corresponding residual sequence; The residual sequence is subjected to time-domain feature extraction to obtain the time-domain evolution features of the current time period; wherein the time-domain evolution features include at least the rate of change feature value and the trend pattern feature.

4. The method according to claim 3, characterized in that, The step of determining the fault type based on the time-domain evolution characteristics of the current time period includes: When the change rate characteristic value is lower than the preset mutation threshold and the trend morphology characteristic shows a monotonically divergent state, the fault type is determined to be a gradual performance degradation fault caused by the cumulative effect of air components. When the rate of change characteristic value is lower than the preset mutation threshold and the trend pattern characteristic shows a non-monotonic divergent state, the fault type is determined to be a transient disturbance fault and filtered out. When the characteristic value of the rate of change is not lower than the preset mutation threshold, the fault type is determined to be a sudden mechanical fault originating from the equipment itself.

5. The method according to claim 4, characterized in that, The operation control of the compressed air energy storage power station according to the fault type includes: When the fault type is the gradual performance degradation fault, adjust the first speed setting value of the low-temperature circulating pump installed on the connecting pipeline between the low-temperature heat storage tank and the compression end heat exchanger, and / or adjust the second speed setting value of the high-temperature circulating pump installed on the connecting pipeline between the high-temperature heat storage tank and the expansion end heat exchanger, and monitor the heat exchange efficiency of the high-temperature circulating pump and / or the low-temperature circulating pump until the heat exchange efficiency reaches the preset deviation range.

6. The method according to claim 5, characterized in that, The operation control of the compressed air energy storage power station according to the fault type includes: When the fault type is the sudden mechanical fault, the first switch valve located at the salt cave inlet section and the second switch valve located at the salt cave outlet section are controlled to perform a rapid closing action, and the power input of the compressor unit or expander unit is cut off simultaneously.

7. The method according to claim 1, characterized in that, The method further includes: Using a parallel time-series feature extraction module of a preset thermodynamic digital twin evolution model, the corresponding time-series feature vector sequence is determined based on multi-dimensional monitoring data of the reference time period; Using the physical constraint decoding module of the preset thermodynamic digital twin evolution model, under the constraint of thermodynamic energy conservation boundary conditions, the theoretical reference value of the current time period is determined according to the time-series feature vector sequence; The preset thermodynamic digital twin evolution model is a deep neural network model based on a heterogeneous dual-stream architecture; the parallel temporal feature extraction module is a module based on a bidirectional long short-term memory network structure with independent dual-channel encoding; and the physical constraint decoding module is a module based on a multilayer perceptron and energy manifold correction structure.

8. A monitoring and control device based on a compressed air energy storage power station, characterized in that, include: The data acquisition module is used to acquire multi-dimensional monitoring data of the compressed air energy storage power station for the current time period, as well as multi-dimensional monitoring data for a reference time period; wherein, the multi-dimensional monitoring data includes thermodynamic state data and air composition data; the air composition data includes at least particulate matter content data and moisture content data; the reference time period includes at least one consecutive compression time period and expansion time period; The reference value determination module is used to determine the theoretical reference value for the current time period by using a preset online simulation model and based on the multidimensional monitoring data of the reference time period. The feature extraction module is used to compare the deviations between the theoretical reference values ​​and the multidimensional monitoring data of the current time period, obtain the deviation comparison results, and extract the temporal evolution features of the current time period based on the deviation comparison results. The operation control module is used to determine the fault type based on the time-domain evolution characteristics of the current time period; and to perform corresponding operation control on the compressed air energy storage power station according to the fault type.

9. An electronic device, characterized in that, It includes a processor and a memory for storing processor-executable instructions, wherein the processor, when executing the instructions, implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images