A method for adaptive regulation of operating temperature of a moisture power generator

Abstract

The application discloses a working temperature self-adaptive adjusting method for a moisture power generation device, and particularly relates to the field of temperature automatic adjustment, and aims to solve the problems that the temperature adjustment in the prior art depends on fixed setting or simple threshold control, cannot reflect the moisture transport state and spatial non-uniformity in the device, and causes power generation performance fluctuation and adjustment lag; the method is characterized in that: evaporation disturbance events are identified by monitoring open-circuit voltages of each subunit and a device surface temperature, voltage recovery and high-frequency internal resistance change characteristics are extracted after the disturbance, an internal resistance recovery characteristic quantity representing humidity gradient reconstruction efficiency is constructed, an in-plane moisture transport anisotropy index is obtained in combination with array space development, a response relationship between the working temperature and the anisotropy index is further established, the optimal working temperature is calculated, and advance adjustment is realized based on a disturbance time interval prediction, so that the self-adaptive optimization control of the working temperature is realized.

Description

Technical Field

[0001] This invention relates to the field of temperature automation control technology, and more specifically, to a method for adaptive adjustment of the operating temperature of a moisture generator. Background Technology

[0002] Moisture-generating devices, a novel type of energy device that utilizes ambient humidity gradients for energy conversion, have been increasingly applied in recent years in scenarios such as self-powered sensing, low-power IoT nodes, and distributed energy harvesting. In practical deployments, these devices are typically installed as thin-film arrays in open or semi-open environments. Their operation is susceptible to temperature fluctuations, airflow, and changes in local evaporation conditions. Especially in applications with significant diurnal temperature variations or frequent changes in the device's own heat generation, the surface moisture distribution and internal humidity gradient exhibit obvious dynamic non-uniformity, leading to fluctuations or even degradation in power generation performance.

[0003] In existing technologies, most solutions regulate the operating temperature of devices by simply setting a fixed temperature setpoint or using a simple threshold control method. This lacks precise perception and dynamic modeling of the internal moisture transport state, making it impossible to effectively control the transient response characteristics during evaporation disturbances. Furthermore, in multi-sub-unit array structures, the moisture transport behavior in different regions exhibits directional differences. Traditional methods struggle to reflect this in-plane anisotropy, leading to a discrepancy between the temperature regulation strategy and the actual moisture migration process, thus affecting overall energy conversion efficiency.

[0004] Therefore, in complex environmental scenarios, how to combine the electrical response and spatial distribution characteristics of the device during operation to construct a characterization index that can reflect the moisture transport state, and thereby achieve adaptive optimization and adjustment of the operating temperature, has become an urgent technical problem to be solved. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an adaptive adjustment method for the operating temperature of a wet gas power generation device to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An adaptive temperature control method for a moisture-generating device includes the following steps: S1. Monitor the open-circuit voltage and surface temperature of each sub-unit of the moisture generator and identify evaporation disturbance events. The sub-unit is a local power generation area formed on the functional film of the moisture generator according to the physical layout coordinates, which is electrically isolated from each other and has independent electrode leads. S2. During the voltage recovery phase after the evaporation disturbance event ends, the open-circuit voltage recovery sequence and high-frequency internal resistance sequence of each sub-unit are collected simultaneously, and the time required for the open-circuit voltage of each sub-unit to recover to the steady-state value before the disturbance is recorded as the apparent recovery time. S3. For each sub-unit, extract the rate of decrease of the high-frequency internal resistance sequence within the apparent recovery time and divide the recovery stage, and calculate the internal resistance recovery feature quantity that characterizes the longitudinal humidity gradient reconstruction efficiency. S4. Expand the internal resistance recovery characteristic of each sub-unit in the array in two dimensions according to the physical layout coordinates, and extract the characteristic difference sequence between adjacent sub-units along the thin film preparation casting direction and the characteristic difference sequence between adjacent sub-units perpendicular to the casting direction. S5. Calculate the ratio of the mean value of the characteristic quantity difference between adjacent sub-units along the film preparation casting direction to the mean value of the characteristic quantity difference between adjacent sub-units perpendicular to the casting direction, and use it as the in-plane moisture transport anisotropy index. S6. Based on historical evaporation disturbance events, establish the correlation between the working environment temperature setpoint and the corresponding in-plane moisture transport anisotropy index, and calculate the optimal working temperature of the wet gas power generation device based on the correlation. S7. Predict the timing of the next disturbance event based on the time interval between historical evaporation disturbance events, and set the operating environment temperature to the optimal operating temperature in advance.

[0008] As a further aspect of the present invention, in step S1, identifying the evaporation disturbance event specifically includes: The open-circuit voltage sequence and the device surface temperature sequence are collected. Within a preset dynamic time window, the difference between the voltage valley value and the voltage value at the beginning of the window is calculated as the voltage disturbance downsurge value. At the same time, the difference between the device surface temperature peak value and the device surface temperature value at the beginning of the window is calculated as the temperature disturbance upsurge value. When the voltage disturbance downsurge exceeds the preset voltage fluctuation limit and the device surface temperature disturbance upsurge exceeds the preset device surface temperature fluctuation limit, an evaporation disturbance event is marked. After an evaporation disturbance event occurs, when the voltage disturbance undershoot is less than the voltage fluctuation limit for a set number of consecutive time points and the device surface temperature disturbance overshoot is less than the device surface temperature fluctuation limit, the end time of the dynamic time window is marked as the end time of the evaporation disturbance event.

[0009] As a further aspect of the present invention, in S2, recording the time required for the open-circuit voltage of each sub-unit to recover to the steady-state value before the disturbance as the apparent recovery time specifically includes: After the evaporation disturbance event ends, the open-circuit voltage amplitude of each sub-unit is collected to form an open-circuit voltage recovery sequence. At the same time, a preset high-frequency sinusoidal current excitation is injected into the sub-unit. The response voltage is measured and the high-frequency internal resistance amplitude of the corresponding sub-unit is calculated to form a high-frequency internal resistance sequence. The frequency of the high-frequency sinusoidal current is higher than the power generation response frequency range of the device. The average open-circuit voltage at a set number of consecutive time points before the evaporation disturbance event is taken as the steady-state reference. Starting from the end of the evaporation disturbance event, each value of the voltage recovery sequence is compared with the steady-state reference point by point. The time from the end of the evaporation disturbance event to the end of the time when the open-circuit voltage first falls into the preset offset range of the steady-state reference is taken as the apparent recovery time of the sub-unit.

[0010] As a further aspect of the present invention, in step S3, calculating the internal resistance recovery characteristic quantity characterizing the longitudinal humidity gradient reconstruction efficiency specifically includes: The subsequence corresponding to the apparent recovery time in the high-frequency internal resistance sequence is smoothed. The second derivative sequence of the smoothed subsequence is taken, and the time corresponding to the first zero-crossing point from the negative value to the positive value in the second derivative sequence is determined as the recovery stage division point. The first average decay rate of the high-frequency internal resistance subsequence before the division point of the recovery stage and the second average decay rate of the high-frequency internal resistance subsequence after the division point of the recovery stage are calculated respectively. The ratio of the second average decay rate to the first average decay rate is used as the internal resistance recovery characteristic of the sub-unit.

[0011] As a further aspect of the present invention, in step S4, extracting the feature quantity difference sequence between adjacent sub-units along the film preparation casting direction and the feature quantity difference sequence between adjacent sub-units perpendicular to the casting direction specifically includes: The internal resistance recovery characteristic of each sub-unit is arranged into a characteristic matrix according to the row and column numbers of each sub-unit in the physical layout coordinate array. The row direction corresponds to the thin film preparation casting direction, and the column direction corresponds to the direction perpendicular to the casting direction. For the feature matrix, the difference in internal resistance recovery feature of adjacent sub-units in the same row is calculated sequentially to construct a feature difference sequence between adjacent sub-units along the flow direction. At the same time, the difference in internal resistance recovery feature of adjacent sub-units in the same column is calculated to construct a feature difference sequence between adjacent sub-units perpendicular to the flow direction.

[0012] As a further aspect of the present invention, in step S5, the in-plane water transport anisotropy index specifically includes: The average difference in the casting direction is obtained by summing the absolute values ​​of each element in the sequence of characteristic quantity differences between adjacent sub-units along the casting direction and dividing by the number of elements in the sequence. At the same time, the average difference in the vertical direction is obtained by summing the absolute values ​​of each element in the sequence of characteristic quantity differences between adjacent sub-units perpendicular to the casting direction and dividing by the number of elements in the sequence. The ratio of the mean difference in the flow direction to the mean difference in the vertical direction is calculated and used as the anisotropy index of in-plane moisture transport.

[0013] As a further aspect of the present invention, in step S6, calculating the optimal operating temperature of the moisture-generating device specifically includes: Extract the in-plane moisture transport anisotropy index corresponding to each disturbance event within the historical monitoring window and the working environment temperature set value recorded at the time of each event to form a sample pair set with the working environment temperature set value as the independent variable and the in-plane moisture transport anisotropy index as the dependent variable. Least square fitting is performed on the sample pair set to obtain the response curve of the in-plane moisture transport anisotropy index as a function of the working environment temperature setpoint. The temperature corresponding to the minimum value of the in-plane moisture transport anisotropy index is found in the response curve and taken as the optimal operating temperature of the moisture power generation device.

[0014] As a further aspect of the present invention, setting the operating environment temperature to the optimal operating temperature in step S7 specifically includes: Record the occurrence times of each historical evaporation disturbance event and calculate adjacent intervals to form an interval sequence. Use the continuous time intervals in the interval sequence as the input feature sequence and input them into the time series-based prediction model for training. The training supervision label is the subsequent real time intervals of the input feature sequence, and the predicted value of the next time interval is output. The estimated occurrence time is obtained by summing the end time of a near-term evaporation disturbance event with the predicted value of the next time interval. Before the estimated occurrence time, the operating environment temperature is adjusted to the optimal operating temperature of the wet gas power generation device.

[0015] The technical effects and advantages of the adaptive adjustment method for the operating temperature of a moisture-generating device according to the present invention are as follows: This invention introduces an evaporation disturbance identification mechanism during the operation of a moisture-generating device and combines the synergistic change characteristics of open-circuit voltage and high-frequency internal resistance to dynamically characterize the device's recovery process after disturbance, achieving a fine depiction of the longitudinal humidity gradient reconstruction behavior. Furthermore, by spatially unfolding the sub-unit array and analyzing the differences between adjacent feature quantities, an in-plane moisture transport anisotropy index is constructed, reflecting the unevenness of moisture distribution on the device surface in different directions. Based on this, a response relationship between operating temperature and the anisotropy index is established, and the optimal operating temperature is determined through an optimization method, transforming temperature regulation from empirical setting to model-driven control based on operating state. Simultaneously, prediction is made by combining historical disturbance event time intervals, allowing temperature regulation to be completed in advance before the next disturbance occurs, improving the foresight and stability of the temperature control strategy. Overall, this method can maintain the coordination of the internal moisture transport state of the device under complex environmental changes, reduce performance fluctuations, improve power generation efficiency and operational stability, and has good engineering applicability and promotional value. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of an adaptive adjustment method for the operating temperature of a moisture-generating device according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Figure 1 The present invention provides an adaptive adjustment method for the operating temperature of a moisture-generating device, which includes the following steps: S1. Monitor the open-circuit voltage and surface temperature of each sub-unit of the moisture generator and identify evaporation disturbance events. The sub-unit is a local power generation area formed on the functional film of the moisture generator according to the physical layout coordinates, which is electrically isolated from each other and has independent electrode leads. S2. During the voltage recovery phase after the evaporation disturbance event ends, the open-circuit voltage recovery sequence and high-frequency internal resistance sequence of each sub-unit are collected simultaneously, and the time required for the open-circuit voltage of each sub-unit to recover to the steady-state value before the disturbance is recorded as the apparent recovery time. S3. For each sub-unit, extract the rate of decrease of the high-frequency internal resistance sequence within the apparent recovery time and divide the recovery stage, and calculate the internal resistance recovery feature quantity that characterizes the longitudinal humidity gradient reconstruction efficiency. S4. Expand the internal resistance recovery characteristic of each sub-unit in the array in two dimensions according to the physical layout coordinates, and extract the characteristic difference sequence between adjacent sub-units along the thin film preparation casting direction and the characteristic difference sequence between adjacent sub-units perpendicular to the casting direction. S5. Calculate the ratio of the mean value of the characteristic quantity difference between adjacent sub-units along the film preparation casting direction to the mean value of the characteristic quantity difference between adjacent sub-units perpendicular to the casting direction, and use it as the in-plane moisture transport anisotropy index. S6. Based on historical evaporation disturbance events, establish the correlation between the working environment temperature setpoint and the corresponding in-plane moisture transport anisotropy index, and calculate the optimal working temperature of the wet gas power generation device based on the correlation. S7. Predict the timing of the next disturbance event based on the time interval between historical evaporation disturbance events, and set the operating environment temperature to the optimal operating temperature in advance.

[0019] In S1, evaporation disturbance events are identified.

[0020] Temperature acquisition points are pre-placed on the surface of the functional thin film of the wet gas power generation device, and an open-circuit voltage acquisition channel is led out from the corresponding electrode of each sub-unit. The acquisition frequency is uniformly set to 100Hz to ensure that the voltage and temperature data are aligned under the same time reference. Based on the continuous data stream acquired in real time, a sliding dynamic time window with a length of 3 seconds is constructed. The time window is updated forward with the current moment as the window tail and a fixed step size. The length of the time window is obtained through multiple sets of experiments. When the time window length is less than 2 seconds, the disturbance characteristics are not obvious, and when it exceeds 5 seconds, the response is lagging. Therefore, 3 seconds is selected as the stable recognition interval. For each time window, the open-circuit voltage sequence is first traversed, and the minimum voltage value within the time window is directly read as the voltage valley value. The voltage value corresponding to the start moment of the time window is read as the window head voltage value. The difference between the two is used as the voltage disturbance downsurge value. This difference reflects the degree of instantaneous voltage collapse during the evaporation enhancement process. Subsequently, the device surface temperature sequence within the same time window was processed. The maximum temperature value in the temperature sequence was read as the temperature peak, and the temperature value at the beginning of the window was read as the temperature reference value. The difference between the two was used as the temperature perturbation overshoot, which reflects the degree of temperature surge caused by environmental or local heat exchange during evaporation. To avoid interference from a single abnormal sampling point on the peak or valley value, a moving average smoothing process with a length of 5 sampling points was performed on the original sequence before calculating the voltage valley and temperature peak. Obvious isolated outliers were removed from the smoothed sequence. Outliers were determined by replacing the neighboring average value when the difference between two adjacent sampling points exceeded three times the historical average value.

[0021] After obtaining the voltage disturbance undershoot and temperature disturbance overshoot, the disturbance event identification stage begins. First, voltage and temperature fluctuation limits are preset. The voltage fluctuation limit is obtained by statistically analyzing the voltage fluctuations of the device under stable operating conditions. Specifically, a time period without significant evaporation disturbances is selected, the average voltage fluctuation amplitude is calculated, and multiplied by 2 to obtain the limit. For example, if the stable voltage fluctuation is 0.02V, the voltage fluctuation limit is set to 0.04V. The temperature fluctuation limit is obtained in a similar manner; for example, if the stable temperature fluctuation is 0.3℃, the temperature fluctuation limit is set to 0.6℃. During the actual judgment process, when the calculated voltage disturbance undershoot within a certain time window exceeds the voltage fluctuation limit, and the corresponding temperature disturbance overshoot within the same time window also exceeds the temperature fluctuation limit, an evaporation disturbance event is immediately determined to have occurred at that time window, and the end time of that time window is recorded as the occurrence time of the disturbance event. To avoid false triggering due to short-term jitter, a constraint requiring simultaneous fulfillment of two conditions is added during the judgment process to ensure physical consistency in the determination of disturbance events. After a disturbance event occurs, continuous rolling detection is performed on subsequent time windows. When the voltage disturbance downsurge amplitude is detected to be less than the voltage fluctuation limit for a consecutive preset number of sampling points, and the temperature disturbance upsurge amplitude is detected to be less than the temperature fluctuation limit for the same number of consecutive sampling points, the evaporation disturbance is determined to have ended, and the termination time of the last time window that meets the conditions is taken as the end time of the disturbance event. The number of consecutive sampling points was experimentally determined to be 10, corresponding to a time length of 0.1 seconds. This value can effectively filter out short-term fluctuations without delaying the end determination time.

[0022] In S2, the time required for the open-circuit voltage of each sub-unit to recover to the steady-state value before the disturbance is recorded as the apparent recovery time.

[0023] The open-circuit voltage of each sub-unit is continuously sampled at a sampling frequency of 200Hz to ensure the rapid changes during the initial recovery phase are captured. The acquired voltage data are arranged in chronological order to form an open-circuit voltage recovery sequence. To avoid abnormal fluctuations in the voltage sequence due to sampling circuit noise or transient interference, a fixed-length sliding average filter of 7 sampling points is applied before the data enters the recovery sequence. The filter window length is determined through comparative experiments. Simultaneously with open-circuit voltage acquisition, a high-frequency sinusoidal current excitation is applied to each sub-unit. The high-frequency current is set to 5kHz, determined through pre-experiments. This frequency is significantly higher than the electrical response frequency range of the device under normal power generation (typically below 100Hz), ensuring that the excitation signal does not interfere with the original power generation process and is only used to extract resistance characteristics. The excitation current amplitude is set to the microampere level, specifically 10μA as the standard value. This amplitude ensures stable acquisition of the response signal without altering the device's humidity distribution. By synchronously measuring the response voltage of each sub-unit under the excitation and calculating the voltage-to-current amplitude ratio, the corresponding high-frequency internal resistance amplitude is obtained, and arranged in chronological order to form a high-frequency internal resistance sequence. To improve the accuracy of internal resistance measurement, the sine wave in each sampling period is sampled for an entire period, and the amplitude is calculated using a peak-to-average method, thereby avoiding errors caused by phase shift. In addition, a short-window smoothing process is also performed on the internal resistance sequence to eliminate random fluctuations in high-frequency measurements.

[0024] After obtaining the open-circuit voltage recovery sequence, the first step is to determine the steady-state reference for determining the termination of recovery. Specifically, the open-circuit voltage data from 20 consecutive sampling points before the evaporation disturbance event are selected as the steady-state interval. This number was obtained through experimental statistics and corresponds to approximately 0.1 seconds at a sampling frequency of 200Hz, which covers the short-term stable interval without introducing long-term drift effects. The average of all voltage values ​​within this interval is used to obtain the steady-state reference voltage. Subsequently, starting from the end of the evaporation disturbance event, the recovery sequence is compared point by point. When the voltage value at a certain moment enters the preset offset range of the steady-state reference, that point is considered to meet the recovery condition. The offset range is determined by statistically analyzing the standard voltage fluctuation amplitude within the steady-state interval, specifically set to 1.5 times the maximum deviation value within the steady-state interval. For example, if the steady-state voltage fluctuation range is ±0.01V, the offset range is set to ±0.015V. To avoid misjudgment caused by instantaneous entry into the offset range, a continuity constraint is introduced during the judgment process. This requires the voltage value to remain within the offset range for 15 consecutive sampling points before that moment is considered the recovery termination point. The end time of the evaporation disturbance event is taken as the starting point, and the recovery termination time that satisfies the above continuity judgment condition is taken as the ending point. The time difference between the two is defined as the apparent recovery time of the subunit. In actual calculations, this time value is obtained by recording the timestamps corresponding to the starting and ending points and directly subtracting them, thus avoiding accumulated errors. For individual abnormal subunits that fail to enter the offset range for an extended period, a maximum recovery time upper limit is set, for example, 5 seconds. If this time is exceeded, the subunit is marked as an abnormal recovery state and separately identified in subsequent processing, excluding it from normal statistics.

[0025] In step S3, the internal resistance recovery characteristic quantity, which characterizes the longitudinal humidity gradient reconstruction efficiency, is calculated.

[0026] After the evaporation disturbance ends, the device surface first undergoes a rapid water film reconstruction process. During this stage, moisture is rapidly redistributed on the surface, and the high-frequency internal resistance shows a rapid decreasing trend, with the curvature exhibiting a continuously concave downward trend. Once the water film is basically restored, the internal humidity gradient begins to dominate the transport process, and moisture diffuses from the interior to the surface. At this point, the rate of change of internal resistance slows down, and the curvature changes from concave to convex, i.e., the second-order change changes from negative to positive. Therefore, this zero-crossing point is precisely the dividing position between the "surface-dominated recovery stage" and the "internal diffusion-dominated stage." Therefore, after obtaining the high-frequency internal resistance sequence corresponding to each sub-unit, the sub-sequence within the apparent recovery time range is first extracted as the internal resistance change data during the recovery period. For this sub-sequence, to eliminate the influence of measurement noise and high-frequency sampling errors on the derivative calculation, a sliding weighted smoothing method with a fixed window length of 9 sampling points is used to process the data. The window length is determined through comparative experiments. When the window length is less than 7, the smoothing is insufficient, and when it is greater than 11, it weakens the true change trend. Therefore, 9 is selected as the stable value. During smoothing, higher weights are assigned to the middle position within the window, while lower weights are assigned to the ends, thus ensuring the trend is maintained while suppressing spikes. After smoothing, a discrete second-order trend calculation is performed on the processed sequence. Specifically, the direction of curvature change is determined point by point through the relationship between three adjacent sampling points, forming a second-order derivative sequence. Subsequently, the sequence is scanned backward from the recovery starting point to find the first transition point that continuously changes from negative to positive and remains positive within a preset number of subsequent sampling points. This number is determined experimentally.

[0027] After determining the recovery phase dividing point, the internal resistance sequence during the recovery period is divided into two subsequences. For the subsequence before the dividing point, the decrease in value between adjacent samples is calculated point by point, and the average of all decreases is taken to obtain the first average decay rate. This rate reflects the overall speed of internal resistance reduction during the rapid reconstruction of the water film in the initial stage after the evaporation disturbance ends. For the subsequence after the dividing point, the decrease in value between adjacent samples is calculated in the same way and the average is taken to obtain the second average decay rate. This rate reflects the slow stage of internal resistance change during the gradual recovery of the internal humidity gradient. Subsequently, the ratio of the second average decay rate to the first average decay rate is calculated to obtain the internal resistance recovery characteristic of this subunit. This ratio has a clear physical meaning: a larger rate in the first stage indicates rapid surface water film recovery, while a smaller rate in the second stage indicates a slower internal humidity gradient recovery process. When the ratio is small, it indicates a significant difference between the two stages, indicating a strong stratification effect in water transport and limited internal diffusion. When the ratio is close to or increases, it indicates a relatively improved recovery speed in the later stage, indicating higher internal humidity gradient reconstruction efficiency and more balanced water transport. Therefore, this characteristic quantity can comprehensively reflect the degree of coordinated recovery of two different physical processes, surface and interior, avoiding the one-sidedness caused by using only a single rate index.

[0028] In step S4, the characteristic difference sequence between adjacent sub-units along the film preparation casting direction and the characteristic difference sequence between adjacent sub-units perpendicular to the casting direction are extracted.

[0029] After calculating the internal resistance recovery characteristic of each subunit, the physical placement of each subunit on the functional thin film is numbered according to the actual structure of the device. Specifically, based on the regular array structure formed during the thin film fabrication process, the film casting direction is used as the row direction, and the direction perpendicular to the casting direction is used as the column direction. Two-dimensional coordinate calibration is performed on all subunits. This calibration process is achieved by recording the spatial position of the electrode leads of each subunit on the substrate. The row number increases sequentially from upstream to downstream according to the casting direction, and the column number increases sequentially from left to right in the horizontal direction. After numbering, the internal resistance recovery characteristic of the corresponding subunit is filled into a two-dimensional matrix according to its row and column positions, forming a characteristic matrix. The row direction corresponds to the main direction in which the material is stretched and deposited during the thin film fabrication process. The microstructure arrangement in this direction exhibits a clear orientation, manifesting as pore structures or moisture-conducting channels extending along the casting direction. Therefore, moisture transport exhibits high continuity in this direction. The column direction corresponds to the horizontal structural direction, where the microstructure distribution is relatively discrete, and there are more obstacles in the moisture transport path. By mapping the internal resistance recovery characteristic to this matrix, not only is the independent recovery capability information of each sub-unit preserved, but it is also embedded into the actual spatial structure, enabling subsequent analysis to directly reflect the spatial distribution differences in moisture transport within the material. This matrix essentially constructs a recovery capability distribution, where high and low values ​​directly correspond to the strength of humidity gradient reconstruction efficiency in different regions, thus providing a basis for identifying directional differences in moisture transport.

[0030] After obtaining the feature matrix, the differences between adjacent sub-units are calculated along two orthogonal directions. First, processing is performed along the flow direction: each row of the matrix is ​​traversed, and within the same row, the internal resistance recovery features of two adjacent sub-units are taken sequentially from left to right according to column number. The difference between the two is calculated, and its absolute value is taken, forming a difference sequence for that row. After processing all rows, the difference sequences are concatenated sequentially to obtain the feature difference sequence between adjacent sub-units along the flow direction. During the calculation, to avoid incomplete data due to missing adjacency relationships in boundary sub-units, only sub-unit pairs with direct adjacency are calculated. Then, processing is performed perpendicular to the flow direction: each column of the matrix is ​​traversed, and within the same column, the feature values ​​of two adjacent sub-units are taken sequentially from top to bottom according to row number. The difference is calculated, and its absolute value is taken, constructing a difference sequence for that column. Finally, these sequences are concatenated to form the difference sequence in the vertical direction. To ensure data comparability between the two directions, the same calculation rules and processing methods are used uniformly when constructing the sequences, and the differences in both directions are expressed in unsigned form. The physical significance of the two difference sequences mentioned above lies in characterizing the spatial gradient change. The difference along the casting direction reflects the degree of continuity of moisture change along the main transport path. When the difference is small and uniformly distributed, it indicates that the transport in that direction is stable and consistent; when the difference fluctuates greatly, it indicates the existence of local transport blockage or imbalance. The difference in the vertical direction reflects the lateral diffusion capability. When the difference is large, it indicates that the lateral humidity gradient difference is significant, and moisture diffusion is restricted. By extracting the difference sequences in these two directions respectively, a quantitative description of the spatial non-uniformity of moisture transport inside the device is achieved.

[0031] In S5, the anisotropy index of in-plane water transport.

[0032] After obtaining the characteristic quantity difference sequences between adjacent sub-units along the film casting direction and the perpendicular direction, statistical processing is performed on the difference sequences in both directions to extract the overall spatial difference level. For the difference sequence in the casting direction, firstly, a consistency check is performed on each difference in the sequence to confirm that they all originate from directly adjacent sub-unit pairs and have undergone absolute value processing, thus ensuring that all values ​​reflect the difference magnitude without including directional information. Subsequently, the sequence is accumulated item by item, and the number of elements participating in the accumulation is recorded. The accumulation process adopts a sequential traversal method to read the difference one by one and add it to the cumulative value. At the same time, obvious abnormal differences are screened during the traversal. When a difference exceeds three times the current mean of the sequence, it is marked as an outlier and replaced with the average of the differences of its left and right adjacent values ​​to avoid the amplified impact of local measurement errors or individual sub-unit anomalies on the overall results. After anomaly processing, the cumulative sum is divided by the number of elements to obtain the mean difference in the casting direction. This mean value characterizes the overall magnitude of the change in water transport capacity between adjacent regions in the main casting direction of the material. The larger the value, the more obvious the spatial imbalance along this direction. Similarly, the same processing procedure is performed on the difference sequence perpendicular to the diffusion direction. Each difference is verified for absolute value and anomaly screening is performed to ensure that the statistically included values ​​reflect the true spatial differences. The same traversal rules are used during the accumulation process, and the mean of adjacent differences is used as a correction value when replacing outliers to maintain data continuity. Finally, the cumulative sum is divided by the number of elements in the difference sequence in that direction to obtain the mean difference in the vertical direction. This mean reflects the degree of change in water transport capacity in the lateral direction. A large value indicates significant obstruction or unevenness in lateral diffusion, while a small value indicates relatively uniform lateral transport.

[0033] After obtaining the average difference between the two directions, the ratio of the two is further calculated as the anisotropy index of in-plane moisture transport. In the calculation, the average difference in the flow direction is used as the numerator, and the average difference in the vertical direction is used as the denominator for the ratio operation. To avoid abnormal results due to the denominator being close to zero, a minimum positive value constraint is set for the average difference in the vertical direction in the actual calculation. When this average is below 0.001, it is uniformly replaced with 0.001. This value is determined through statistical analysis of experimental data and is far below the normal difference level, so it will not significantly deviate from the actual ratio. The anisotropy index obtained through this ratio definition is a dimensionless quantity, and its value directly reflects the relative relationship of the degree of difference between the two directions: when the index is close to 1, it indicates that the difference level in the two directions is comparable, indicating that in-plane moisture transport tends to be uniform; when the index is significantly greater than 1, it indicates that the difference in the flow direction is greater, indicating that there is obvious transport unevenness in that direction; when the index is close to 0, it indicates that the difference in the vertical direction is greater, and lateral transport is restricted.

[0034] In step S6, the optimal operating temperature of the moisture generator is calculated.

[0035] A historical monitoring window is maintained in a continuously rolling manner over time to store data related to recent evaporation disturbance events. The length of the monitoring window is determined experimentally to be 30 disturbance events. This number is sufficient to cover various temperature conditions in actual operation without introducing premature historical data that could interfere with the current state. After a disturbance event ends, the in-plane moisture transport anisotropy index corresponding to the event is paired with the ambient temperature setpoint recorded at the time of the event to form a set of sample data, which is then added to the sample pair set sequentially. During the sample pair construction process, the ambient temperature setpoint is checked for consistency to ensure that it corresponds to the actual control setpoint at the time of the disturbance event, rather than a transitional value during the adjustment process. Specifically, by recording the setting update log of the temperature control execution unit, the most recent stable setpoint is read as the temperature corresponding to each disturbance event, thereby ensuring a clear correspondence between temperature data and anisotropy index. To improve the uniformity of sample distribution, when the temperature distribution within the monitoring window is concentrated in a small range, historical data is sampled in segments, and representative data in different temperature ranges are retained first to avoid situations where local data is too dense while other ranges are sparse during the fitting process.

[0036] After constructing the sample set, a fitting process is performed to establish the response relationship between temperature and the anisotropy index. Specifically, a least-squares fitting method is used, which involves traversing candidate functions and calculating the fitting error, selecting the function with the smallest error as the response curve representation. In practice, a quadratic polynomial function is preferred for fitting, as it can describe the trend of change with a single extreme point and exhibits high stability during calculation. During fitting, all sample points are weighted uniformly, and previously marked outliers are excluded from the fitting calculation, ensuring that the curve shape is not affected by outliers. After fitting, the smoothness of the obtained response curve is checked to ensure that there are no abrupt changes or multiple extreme points within the sample range. If multiple extreme points are detected, the fitting is re-executed with increased sample screening intensity until the curve exhibits a single extreme value characteristic. After obtaining the response curve, a traversal search is performed on the curve to determine the temperature point that minimizes the in-plane moisture transport anisotropy index. Specifically, a scan is performed within the sample temperature range with a fixed step size of 0.5℃, calculating the corresponding index value point by point and recording the temperature position of the minimum value. To improve accuracy, after determining the coarse minimum value, a further refined search is performed in the vicinity of this temperature using smaller step sizes of 0.1℃, resulting in a more precise optimal operating temperature. To avoid misjudgments due to local fluctuations, when the difference between the minimum point and its adjacent temperature points is less than a preset tolerance range, the average temperature within that interval is taken as the final result. The final optimal operating temperature is directly used as the target for subsequent temperature control adjustments; this temperature corresponds to the state where the device's moisture transport is most balanced under the current environmental conditions.

[0037] In step S7, the ambient temperature is set to the optimal operating temperature.

[0038] The occurrence time of each evaporation disturbance event is recorded, specifically by directly collecting the timestamps output from the aforementioned disturbance identification steps and storing them in a historical record table in chronological order. Subsequently, the difference between the occurrence times of two adjacent disturbance events is calculated to obtain a continuous time interval sequence. When constructing the prediction model input data, fixed-length continuous time interval segments are used as input feature sequences. The segment length is set to 8, which was obtained through statistical analysis of multiple sets of running data. When the length is less than 5, the model cannot effectively learn periodic variation features; when it exceeds 12, redundant information is introduced and training efficiency is reduced. Therefore, 8 is selected as a balance value. The supervision label corresponding to each input sequence is the real time interval immediately following that sequence, thus forming training samples. The prediction model adopts a time-series-based Long Short-Term Memory (LSTM) structure, specifically including an input layer, a LSM unit layer, and an output layer. The number of LSM units is set to 16, which can effectively capture the short-term variation trend and potential periodic patterns of disturbance intervals. During model training, mean squared error was used as the loss function, with 200 training epochs and a fixed learning rate of 0.01. The prediction error was gradually converged by updating the model parameters epoch by epoch. After training, the model was validated to ensure that its prediction error on historical data remained consistently below a preset error range, for example, with the average error controlled within 0.2 seconds, thus guaranteeing the model's practical application capability. Through this training process, the model can output the predicted value for the next time interval based on the recent trend of perturbation time interval changes.

[0039] The system continuously acquires the end time of the most recent evaporation disturbance event and extracts the preceding eight consecutive time intervals as the current input feature sequence. This sequence is then input into a trained time series prediction model to obtain the predicted value for the next time interval. This predicted value represents the expected time span of the next evaporation disturbance event. This time span is then superimposed with the end time of the most recent disturbance event to obtain the estimated occurrence time of the next disturbance event. To ensure the stability of the prediction results, boundary constraints are applied to the model output values ​​in practical use. When the predicted value is less than 80% of the historical minimum interval or greater than 120% of the historical maximum interval, it is limited to this range to avoid unreasonable outputs from the model in extreme cases. After obtaining the estimated occurrence time, a temperature adjustment lead time is set. This lead time, determined experimentally to be 2 seconds, is sufficient for the ambient temperature to transition from the current value to the target optimal operating temperature, while avoiding premature adjustment that could increase energy consumption. During execution, when the current time approaches 2 seconds before the estimated occurrence time, the temperature adjustment mechanism is activated to gradually adjust the operating ambient temperature to the optimal operating temperature determined in the aforementioned steps. The adjustment process employs a linear incremental approach, ensuring a smooth temperature change to the target value within one second, thereby preventing secondary disturbances to the device's operating state caused by sudden temperature changes. To guarantee adjustment accuracy, the device surface temperature is continuously monitored after temperature adjustment. If the actual temperature deviates from the target value by more than ±0.3℃, fine-tuning is immediately performed to stabilize the temperature within the target range.

[0040] Through the above process, temperature pre-regulation is achieved before evaporation disturbances occur, ensuring that the device is already at its optimal operating temperature when the disturbance arrives. The entire process is based on historical time interval sequences, using a trained predictive model to anticipate the disturbance time in advance. Combined with a temperature regulation strategy, this transforms temperature control from a passive response to an active regulation, ensuring the continuity and stability of the control process.

[0041] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0042] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0043] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0045] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0046] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0047] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0049] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for adaptive adjustment of operating temperature for a moisture-generating device, characterized in that, Includes the following steps: S1. Monitor the open-circuit voltage and surface temperature of each sub-unit of the moisture generator and identify evaporation disturbance events. The sub-unit is a local power generation area formed on the functional film of the moisture generator according to the physical layout coordinates, which is electrically isolated from each other and has independent electrode leads. S2. During the voltage recovery phase after the evaporation disturbance event ends, the open-circuit voltage recovery sequence and high-frequency internal resistance sequence of each sub-unit are collected simultaneously, and the time required for the open-circuit voltage of each sub-unit to recover to the steady-state value before the disturbance is recorded as the apparent recovery time. S3. For each sub-unit, extract the rate of decrease of the high-frequency internal resistance sequence within the apparent recovery time and divide the recovery stage, and calculate the internal resistance recovery feature quantity that characterizes the longitudinal humidity gradient reconstruction efficiency. S4. Expand the internal resistance recovery characteristic of each sub-unit in the array in two dimensions according to the physical layout coordinates, and extract the characteristic difference sequence between adjacent sub-units along the thin film preparation casting direction and the characteristic difference sequence between adjacent sub-units perpendicular to the casting direction. S5. Calculate the ratio of the mean value of the characteristic quantity difference between adjacent sub-units along the film preparation casting direction to the mean value of the characteristic quantity difference between adjacent sub-units perpendicular to the casting direction, and use it as the in-plane moisture transport anisotropy index. S6. Based on historical evaporation disturbance events, establish the correlation between the working environment temperature setpoint and the corresponding in-plane moisture transport anisotropy index, and calculate the optimal working temperature of the wet gas power generation device based on the correlation. S7. Predict the timing of the next disturbance event based on the time interval between historical evaporation disturbance events, and set the operating environment temperature to the optimal operating temperature in advance.

2. The adaptive temperature adjustment method for a wet gas power generation device according to claim 1, characterized in that, In step S1, identifying evaporation disturbance events specifically includes: The open-circuit voltage sequence and the device surface temperature sequence are collected. Within a preset dynamic time window, the difference between the voltage valley value and the voltage value at the beginning of the window is calculated as the voltage disturbance downsurge value. At the same time, the difference between the device surface temperature peak value and the device surface temperature value at the beginning of the window is calculated as the temperature disturbance upsurge value. When the voltage disturbance downsurge exceeds the preset voltage fluctuation limit and the device surface temperature disturbance upsurge exceeds the preset device surface temperature fluctuation limit, an evaporation disturbance event is marked. After an evaporation disturbance event occurs, when the voltage disturbance undershoot is less than the voltage fluctuation limit for a set number of consecutive time points and the device surface temperature disturbance overshoot is less than the device surface temperature fluctuation limit, the end time of the dynamic time window is marked as the end time of the evaporation disturbance event.

3. The adaptive temperature adjustment method for a wet gas power generation device according to claim 1, characterized in that, In step S2, the time required for the open-circuit voltage of each sub-unit to recover to its pre-disturbance steady-state value is recorded as the apparent recovery time. Specifically, this includes: After the evaporation disturbance event ends, the open-circuit voltage amplitude of each sub-unit is collected to form an open-circuit voltage recovery sequence. At the same time, a preset high-frequency sinusoidal current excitation is injected into the sub-unit. The response voltage is measured and the high-frequency internal resistance amplitude of the corresponding sub-unit is calculated to form a high-frequency internal resistance sequence. The frequency of the high-frequency sinusoidal current is higher than the power generation response frequency range of the device. The average open-circuit voltage at a set number of consecutive time points before the evaporation disturbance event is taken as the steady-state reference. Starting from the end of the evaporation disturbance event, each value of the voltage recovery sequence is compared with the steady-state reference point by point. The time from the end of the evaporation disturbance event to the end of the time when the open-circuit voltage first falls into the preset offset range of the steady-state reference is taken as the apparent recovery time of the sub-unit.

4. The adaptive temperature control method for a wet gas power generation device according to claim 1, characterized in that, In step S3, the calculation of the internal resistance recovery characteristic quantity characterizing the longitudinal humidity gradient reconstruction efficiency specifically includes: The subsequence corresponding to the apparent recovery time in the high-frequency internal resistance sequence is smoothed. The second derivative sequence of the smoothed subsequence is taken, and the time corresponding to the first zero-crossing point from the negative value to the positive value in the second derivative sequence is determined as the recovery stage division point. The first average decay rate of the high-frequency internal resistance subsequence before the division point of the recovery stage and the second average decay rate of the high-frequency internal resistance subsequence after the division point of the recovery stage are calculated respectively. The ratio of the second average decay rate to the first average decay rate is used as the internal resistance recovery characteristic of the sub-unit.

5. The adaptive temperature adjustment method for a wet gas power generation device according to claim 1, characterized in that, In step S4, extracting the feature difference sequence between adjacent sub-units along the film preparation casting direction and the feature difference sequence between adjacent sub-units perpendicular to the casting direction specifically includes: The internal resistance recovery characteristic of each sub-unit is arranged into a characteristic matrix according to the row and column numbers of each sub-unit in the physical layout coordinate array. The row direction corresponds to the thin film preparation casting direction, and the column direction corresponds to the direction perpendicular to the casting direction. For the feature matrix, the difference in internal resistance recovery feature of adjacent sub-units in the same row is calculated sequentially to construct a feature difference sequence between adjacent sub-units along the flow direction. At the same time, the difference in internal resistance recovery feature of adjacent sub-units in the same column is calculated to construct a feature difference sequence between adjacent sub-units perpendicular to the flow direction.

6. The adaptive temperature control method for a wet gas power generation device according to claim 1, characterized in that, In S5, the in-plane water transport anisotropy index specifically includes: The average difference in the casting direction is obtained by summing the absolute values ​​of each element in the sequence of characteristic quantity differences between adjacent sub-units along the casting direction and dividing by the number of elements in the sequence. At the same time, the average difference in the vertical direction is obtained by summing the absolute values ​​of each element in the sequence of characteristic quantity differences between adjacent sub-units perpendicular to the casting direction and dividing by the number of elements in the sequence. The ratio of the mean difference in the flow direction to the mean difference in the vertical direction is calculated and used as the anisotropy index of in-plane moisture transport.

7. The adaptive temperature adjustment method for a wet gas power generation device according to claim 1, characterized in that, In step S6, calculating the optimal operating temperature of the moisture generator specifically includes: Extract the in-plane moisture transport anisotropy index corresponding to each disturbance event within the historical monitoring window and the working environment temperature set value recorded at the time of each event to form a sample pair set with the working environment temperature set value as the independent variable and the in-plane moisture transport anisotropy index as the dependent variable. Least square fitting is performed on the sample pair set to obtain the response curve of the in-plane moisture transport anisotropy index as a function of the working environment temperature setpoint. The temperature corresponding to the minimum value of the in-plane moisture transport anisotropy index is found in the response curve and taken as the optimal operating temperature of the moisture power generation device.

8. The adaptive temperature control method for a wet gas power generation device according to claim 1, characterized in that, In step S7, setting the operating ambient temperature to the optimal operating temperature specifically includes: Record the occurrence times of each historical evaporation disturbance event and calculate adjacent intervals to form an interval sequence. Use the continuous time intervals in the interval sequence as the input feature sequence and input them into the time series-based prediction model for training. The training supervision label is the subsequent real time intervals of the input feature sequence, and the predicted value of the next time interval is output. The estimated occurrence time is obtained by summing the end time of a near-term evaporation disturbance event with the predicted value of the next time interval. Before the estimated occurrence time, the operating environment temperature is adjusted to the optimal operating temperature of the wet gas power generation device.

Images