New energy power supply low-power satellite data collection terminal method and system

Abstract

The application relates to the technical field of data collection, and discloses a new energy power supply low-power satellite data collection terminal method and system, wherein the method comprises the following steps: monitoring residual power data of a communication terminal, and combining environmental parameters to calculate energy budget prediction data; performing compression on collected data based on the energy budget prediction data to obtain target compression data; creating a scheduling scheme of the target compression data; distributing the target compression data to a delay queue and matching an effective communication window according to the scheduling scheme to obtain a transmission queue and select a transmission mode to send target compression data in the transmission queue, thereby obtaining a satellite transmission result; and the application can automatically select an optimal working mode combination according to a power state, and maximally prolongs communication time under energy constraints.

Description

Technical Field

[0001] This invention relates to the field of data acquisition technology, and in particular to a low-power satellite data acquisition terminal method and system powered by new energy sources. Background Technology

[0002] Low-power satellite data acquisition terminals powered by renewable energy sources are used in environments without power, such as ocean buoys, field monitoring stations, and environmental monitoring in remote areas. These terminals typically employ a hybrid power supply of solar and wind power, using satellite communication technology to achieve remote data transmission. However, renewable energy power supply systems are affected by various factors such as weather conditions, seasonal changes, and geographical location, resulting in significant fluctuations and uncertainties in power output and extremely unstable energy supply to the data acquisition terminals. Traditional methods usually employ fixed data acquisition frequencies and uniform data processing strategies, failing to dynamically adjust according to actual energy conditions, and are prone to insufficient power supply under extreme weather conditions such as continuous rain or no wind. Summary of the Invention

[0003] This invention provides a low-power satellite data acquisition terminal method and system powered by new energy sources. This invention can automatically select the optimal combination of working modes according to the power status, thereby maximizing the extension of communication time under energy constraints.

[0004] In a first aspect, the present invention provides a method for a low-power satellite data acquisition terminal powered by a new energy source, the method comprising: Monitor the remaining battery power of communication terminals and calculate energy budget forecast data in combination with environmental parameters; Based on the energy budget prediction data, the collected data is compressed to obtain the target compressed data; Create a scheduling scheme for the target compressed data; According to the scheduling scheme, the target compressed data is allocated to the delay queue and matched with a valid communication window to obtain a transmission queue. The transmission mode is selected to send the target compressed data in the transmission queue, and the satellite transmission result is obtained.

[0005] In conjunction with the first aspect, in a first implementation of the first aspect of the present invention, the monitoring of the remaining battery power data of the communication terminal and the calculation of energy budget prediction data in conjunction with environmental parameters include: Acquire instantaneous current and temperature data of the battery in the communication terminal; Based on the battery temperature data and the number of charge-discharge cycles, the battery internal resistance is compensated and corrected to obtain temperature-corrected SOC data. A battery health state matrix is ​​constructed based on the instantaneous current data, the battery temperature data, and the temperature-corrected SOC data; The remaining available energy is calculated based on the battery health status matrix to obtain the remaining power data; Based on the remaining power data and environmental parameters, the power supply of new energy sources is predicted to obtain energy budget prediction data.

[0006] In conjunction with the first aspect, in a second implementation of the first aspect of the present invention, the step of predicting the power supply of new energy sources based on the remaining power data and environmental parameters to obtain energy budget prediction data includes: The environmental parameters of the communication terminal are obtained, and light and wind speed data are extracted from the environmental parameters. Temporal feature extraction and attention weight calculation are performed on the illumination data to obtain solar power prediction results; The wind speed data is modeled by probability distribution and mapped by nonlinear power curve to obtain wind power prediction results; By combining the remaining power data, energy balance calculations are performed on the solar power prediction results and the wind power prediction results to obtain energy budget prediction data.

[0007] In conjunction with the first aspect, in a third implementation of the first aspect of the present invention, the step of performing probability distribution modeling and nonlinear power curve mapping on the wind speed data to obtain wind power prediction results includes: Wind speed, gust coefficient, atmospheric pressure, and air density are extracted from the wind speed data, and a wind energy prediction input vector is constructed. A wind turbine power curve model is established based on the relationship between air density and wind speed in the wind energy prediction input vector. Gradient vanishing compensation is performed on the wind energy prediction input vector to obtain the wind speed change feature vector; An attention mechanism is applied to the wind speed change feature vector to identify the moment of sudden wind speed change, and power is calculated by combining it with the wind turbine power curve model to obtain the wind power prediction result.

[0008] In conjunction with the first aspect, in a fourth implementation of the first aspect of the present invention, the step of compressing the collected data based on the energy budget prediction data to obtain the target compressed data includes: Based on the energy budget forecast data, the collected data is prioritized to obtain emergency alarm data, abnormal event data, trend monitoring data, periodic routine data, and redundant background data. The emergency alarm data is subjected to lossless compression with CRC32 checksum to obtain the first compressed data; the abnormal event data is subjected to adaptive arithmetic coding to obtain the second compressed data; the trend monitoring data is subjected to piecewise linear fitting to obtain the third compressed data; the periodic regular data is subjected to incremental coding to obtain the fourth compressed data; and the redundant background data is subjected to spatiotemporal aggregation processing to obtain the fifth compressed data. The first compressed data, the second compressed data, the third compressed data, the fourth compressed data, and the fifth compressed data are respectively encapsulated to obtain the identified compressed data; The compression level of the labeled compressed data is adjusted based on the energy budget forecast data to obtain the target compressed data.

[0009] In conjunction with the first aspect, in a fifth implementation of the first aspect of the present invention, the scheduling scheme for creating the target compressed data includes: Extract data priority weight, queue delay time, and satellite transit time parameters from the target compressed data, and use the data priority weight, queue delay time, and satellite transit time parameters as input parameters; The weighting coefficients in the utility function are adjusted based on the remaining power data to obtain the power perception weighting coefficients. The input parameters and the power perception weight coefficient are input into the utility function to calculate the utility value, and the utility value of the transmission task is obtained. Based on the utility value of the transmission task, a task selection is performed to obtain a scheduling scheme.

[0010] In conjunction with the first aspect, in a sixth implementation of the first aspect of the present invention, the step of selecting tasks based on the transmission task utility value to obtain a scheduling scheme includes: Construct a utility sorting queue by sorting the utility values ​​of the transmission tasks in descending order; The transmission energy quota is calculated based on the remaining power data and the predicted average available power. The transmission tasks are selected one by one from the utility sorting queue and the transmission power consumption is accumulated until the transmission energy quota is reached, thus obtaining a set of transmission tasks. The utility values ​​of the transmission tasks in the transmission task set are recalculated and queue adjustments are performed to obtain a scheduling scheme.

[0011] In conjunction with the first aspect, in the seventh implementation of the first aspect of the present invention, the step of allocating the target compressed data to a delay queue and matching a valid communication window according to the scheduling scheme to obtain a transmission queue and selecting a transmission mode to send the target compressed data in the transmission queue, thereby obtaining a satellite transmission result, includes: A delay queue is constructed based on the remaining battery power data, and a dynamic timeout threshold is set for the delay queue; Based on the scheduling scheme, the target compressed data is placed into the delay queue according to priority, and the satellite transit time and elevation angle parameters are calculated through the dual-line element set data to obtain the effective communication window; Before the start of the effective communication window, select the transmission task in the delay queue to obtain the transmission queue; Based on the remaining battery power data, a transmission mode is selected and the target compressed data in the transmission queue is sent to obtain the satellite transmission result.

[0012] In conjunction with the first aspect, in the eighth implementation of the first aspect of the present invention, the step of selecting a transmission mode based on the remaining battery power data and sending the target compressed data in the transmission queue to obtain the satellite transmission result includes: Based on the remaining power data and energy harvesting power, determine whether to activate the shadow transmission mode and obtain a mode switching command; According to the transmission mode corresponding to the mode switching instruction, the trend monitoring data, periodic regular data and redundant background data in the transmission queue are converted into BeiDou short message format to obtain low-power transmission data. The transmission power consumption of the low-power transmission data and the conventional satellite communication data are calculated respectively, and the optimal transmission path is selected to obtain the transmission path scheme; Data transmission is performed using the aforementioned transmission path scheme, and the transmission success rate is monitored to obtain the satellite transmission results.

[0013] Secondly, the present invention provides a low-power satellite data acquisition terminal system powered by a new energy source, the low-power satellite data acquisition terminal system powered by a new energy source comprising: The monitoring module is used to monitor the remaining battery power of the communication terminal and calculate the energy budget forecast data in combination with environmental parameters. The compression module is used to compress the collected data based on the energy budget prediction data to obtain the target compressed data; A creation module is used to create a scheduling scheme for the target compressed data; The transmission module is used to allocate the target compressed data to a delay queue and match a valid communication window according to the scheduling scheme, obtain a transmission queue, select a transmission mode to send the target compressed data in the transmission queue, and obtain the satellite transmission result.

[0014] The technical solution provided by this invention combines temperature resistance compensation and multi-sensor fusion to achieve high-precision monitoring of remaining power, overcoming the large errors of traditional voltage detection methods. It predicts the output power of solar and wind power, automatically learning the periodicity of light intensity and the randomness of wind speed. A five-level data priority classification system is established, and specialized compression algorithms are used for data of different importance, maximizing transmission efficiency while ensuring the quality of critical data and avoiding resource waste caused by traditional uniform compression methods. Based on the energy marginal utility function, the value of data transmission is accurately quantified, comprehensively considering the dynamic balance between data importance, transmission delay, and energy consumption, overcoming the limitations of fixed scheduling strategies. Through dual-line element set orbit prediction and multi-level delay-tolerant queue management, accurate calculation of satellite transit times and full utilization of communication windows are achieved, solving the energy waste problem caused by blind transmission. In extremely low power conditions, the BeiDou short message backup transmission mechanism is activated, significantly reducing power consumption while ensuring reliable transmission of critical data. A refined energy allocation strategy and global optimization control mechanism are established, automatically selecting the optimal combination of working modes based on power status, maximizing communication time under energy constraints.

[0015] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of an embodiment of the low-power satellite data acquisition terminal method powered by new energy sources in this invention. Figure 2 This is a schematic diagram of one embodiment of a low-power satellite data acquisition terminal system powered by new energy sources according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions 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, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The terms "comprising" and "having," and any variations thereof, used in the embodiments of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] To facilitate understanding of this embodiment, a detailed description of a low-power satellite data acquisition terminal method powered by new energy sources, as disclosed in this embodiment of the invention, will be provided first. For example... Figure 1 As shown, the low-power satellite data acquisition terminal method powered by new energy sources includes the following steps: 101. Monitor the remaining battery power of communication terminals and calculate energy budget forecast data in combination with environmental parameters; Specifically, instantaneous current and temperature data of the battery are acquired through sensors integrated within the communication terminal. A battery internal resistance compensation model based on the combined effects of temperature and aging factors is introduced. The current temperature data and historical charge-discharge cycle count of the battery are substituted into the battery internal resistance calculation formula. A correction factor is used to dynamically correct the trend of internal resistance change with temperature drift and aging resistance increase, correcting the SOC value obtained based on the current integration method to form temperature-corrected SOC data. The temperature-corrected SOC data, together with current, voltage, temperature, and SOH state, construct a five-dimensional battery health state matrix, reflecting the real-time health level of the battery system under complex operating conditions. By multiplying the current capacity, voltage, temperature correction factor, load correction factor, and discharge efficiency parameters extracted from the battery health state matrix, the remaining usable energy of the communication terminal under the current energy consumption mode is calculated, and the remaining power data is obtained based on this. Based on the remaining power data, environmental parameters such as historical light intensity sequences, cloud cover, relative humidity, ambient temperature, and seasonal factors are input into the LSTM-Attention network for photovoltaic power prediction. At the same time, historical wind speed, gust coefficient, atmospheric pressure, and air density data are input into the residual connection network for wind power modeling. Combined with the current load power model, the energy balance curve of the prediction system over 72 hours is calculated to obtain the time-by-time difference of photovoltaic power generation, wind power generation, and load energy consumption, and the energy balance prediction result is output.

[0021] 102. Based on the energy budget forecast data, compress the collected data to obtain the target compressed data; Specifically, energy budget forecast data is used as the basis for compression strategy regulation. By constructing a priority determination model driven by both energy constraints and data importance, the collected data is subjected to content identification and dynamic classification. The collected data is divided into five levels: emergency alarm data, abnormal event data, trend monitoring data, periodic routine data, and redundant background data. Emergency alarm data corresponds to signals with high suddenness and high severity, such as sensor over-limit or equipment failure information. Abnormal event data corresponds to system disturbances with medium frequency but obvious characteristics. Trend monitoring data is used to capture long-term evolutionary change trajectories. Periodic routine data is mostly highly redundant timed record information. Redundant background data includes low-value data such as duplicate samples and logs. Based on the priority characteristics of each type of data, the most suitable compression algorithm is matched for processing. Emergency alarm data uses a lossless compression algorithm with CRC32 check mechanism to ensure that no content is lost during transmission, resulting in the first compressed data. Abnormal event data uses an adaptive arithmetic coding method to dynamically model the probabilities of different symbols and perform efficient entropy coding, resulting in the second compressed data. Trend monitoring data retains the main trend structure through piecewise linear fitting and achieves balanced compression with a fitting error control mechanism, forming the third compressed data. Periodic regular data uses incremental coding technology to record only the difference with the data at the previous moment and aggregates and transmits it under threshold control, generating the fourth compressed data. Redundant background data is input into the spatiotemporal aggregation algorithm processing unit to merge and compress similar data within the same time window, extract their statistical feature values, and generate the fifth compressed data. The first, second, third, fourth, and fifth compressed data are encapsulated separately. Each type of compressed data is appended with an identifier header containing a priority tag, timestamp, compression ratio, and checksum to form an identifier compressed data. The compression level of each type of identifier compressed data is dynamically adjusted based on the energy budget forecast results. When an energy surplus is predicted, the compression intensity is reduced to improve data fidelity, while when a deficit is predicted, the compression ratio is increased to control the transmission load, and the target compressed data is output.

[0022] 103. Create a scheduling scheme for the target compressed data; Specifically, key attribute parameters for scheduling decisions are extracted from the target compressed data, including data priority weights, the cumulative delay time of the data in the queue, and the corresponding satellite transit time. Data priority weights reflect the importance level of the data, delay time reflects the waiting time of the target compressed data in the cache, and satellite transit time indicates the time distance to the next available transmission window. These three types of parameters constitute the input variables for scheduling decisions. Simultaneously, the current remaining power data is obtained from the energy sensing module, and the weight coefficients in the utility function are dynamically adjusted according to the remaining power's operating mode range (e.g., sufficient, energy-saving, warning, or emergency). By modifying the parameters α, β, and γ in the function, different degrees of emphasis are applied to data integrity, delay tolerance, and satellite time sensitivity, forming power sensing weight coefficients that match the current energy state. The extracted data priority weights, queue delay time, and satellite transit time, along with the power sensing weight coefficients, are input into the utility function model for calculation. The scheduling value of each compressed data packet under current energy and delay constraints is calculated. The utility function comprehensively considers the current task's priority in resource competition, the degree of delay penalty, and the urgency of the satellite link. All transmission tasks are sorted according to their utility value, and the task with the highest utility value is selected and added to the current transmission task set based on the available energy budget and satellite window capacity, thus generating a scheduling scheme.

[0023] 104. According to the scheduling scheme, the target compressed data is allocated to the delay queue and matched with the valid communication window to obtain the transmission queue. The transmission mode is selected to send the target compressed data in the transmission queue to obtain the satellite transmission result.

[0024] Specifically, the delay queue structure is constructed based on the current remaining power data, and the dynamic timeout threshold for each priority queue is adaptively set according to the power sufficiency. Under high power conditions, each queue maintains a longer waiting time to obtain a better transmission window. When power is scarce, the maximum dwell time of low-priority queues is automatically shortened to ensure priority processing of critical data. The remaining power is mapped to the timeout time of each queue using an exponential function, making the timeout duration non-linearly sensitive to power changes. Based on the priority level of each target compressed data in the scheduling scheme, the data is sequentially allocated to the corresponding delay queue, forming a hierarchical buffer structure. Simultaneously, the system accesses TLE data containing orbital parameters and uses the SGP4 orbital analysis algorithm to calculate satellite positions, obtaining the transit time and maximum elevation angle of each satellite in real time for the next 24 hours. Time segments with elevation angles greater than 10 degrees and link quality functions exceeding the threshold are identified as effective communication windows, determining the available time periods for communication scheduling. During a pre-processing period before each communication window arrives, all tasks in the delay queue are actively scanned. A greedy bin packing algorithm is used to match the compressed data size of each task with the remaining window capacity, and the task utility value is combined for sorting and optimization to generate the corresponding transmission queue for the window. After entering the effective period of the communication window, the most suitable satellite transmission mode is selected based on the current remaining power data and power budget. When the power is above the threshold and the power is sufficient, the Tiantong or Iridium satellite mode with larger bandwidth is given priority. When the power is below the warning threshold or the power budget is insufficient, the system switches to the low-power Beidou short message transmission mode and uses a fixed-point compression format to reduce the transmission load. Data packets in the transmission queue are uploaded to the satellite link one by one, and information such as power consumption, latency, and response results are recorded during the transmission process to form satellite transmission result feedback.

[0025] In one specific embodiment, the process of performing step 101 may specifically include the following steps: Acquire instantaneous current and temperature data of the battery in the communication terminal; The battery internal resistance is compensated and corrected based on battery temperature data and charge / discharge cycle count to obtain temperature-corrected SOC data. A battery health state matrix is ​​constructed based on instantaneous current data, battery temperature data, and temperature-corrected SOC data; The remaining available energy is calculated based on the battery health state matrix to obtain the remaining power data; Based on the remaining electricity data and environmental parameters, the power supply of new energy sources is predicted to obtain energy balance prediction data.

[0026] Specifically, a coulomb counter and a temperature sensor are deployed in the terminal battery management system to capture the instantaneous current value and current internal temperature data of the battery during operation. The current sampling frequency is controlled at the millisecond level to reflect dynamic load response characteristics, and the temperature sampling accuracy is controlled within ±0.5℃ to meet the input requirements of the internal resistance compensation model. The collected temperature data and the recorded historical charge-discharge cycle count are input into the battery internal resistance modeling unit. By establishing a temperature-aging coupled compensation function, the compensated dynamic internal resistance of the battery is output in real time under bivariate conditions. The corrected internal resistance data is combined with the instantaneous current value to improve the accuracy of the coulomb integral calculation, and the SOC value is dynamically offset by combining it with the temperature correction model to output temperature-corrected SOC data that is closer to the actual battery state of charge. A five-dimensional battery health state matrix is ​​constructed, including temperature-corrected SOC, current, voltage, temperature, and SOH (state of health). The current battery capacity, voltage, discharge efficiency, temperature correction coefficient, and load correction coefficient are input into the energy calculation module. The product model is used to obtain the available energy value under the current real conditions, and the energy value is divided by the predicted average load power to estimate the remaining working time, outputting quantified remaining energy data. Using historical sensor data from both the photovoltaic (PV) and wind power subsystems as input, a PV power prediction network is constructed with historical irradiance sequences, cloud cover, relative humidity, ambient temperature, and seasonal factors as inputs. Simultaneously, a wind power prediction vector is constructed with historical wind speed sequences, gust factors, atmospheric pressure, and air density as inputs. Both types of data are then input into a bidirectional LSTM-Attention network and a residual network architecture for hourly power generation prediction over 72 hours. The predicted load power, PV power, and wind power are input into an energy budget function, and their hourly trends are statistically analyzed to establish an energy budget prediction curve. The output covers the distribution of energy surplus and deficit periods over the future timeframe.

[0027] In one specific embodiment, the process of performing the step of predicting the power supply of new energy sources based on the remaining power data and environmental parameters to obtain energy budget prediction data can specifically include the following steps: Obtain the environmental parameters of the communication terminal and extract light and wind speed data from the environmental parameters; Temporal features are extracted and attention weights are calculated from the illumination data to obtain the solar power prediction results; By performing probability distribution modeling and nonlinear power curve mapping on wind speed data, wind power prediction results are obtained. By combining the remaining electricity data, energy balance calculations are performed on the solar power prediction results and the wind power prediction results to obtain energy budget prediction data.

[0028] Specifically, the environmental perception module collects real-time information about the external natural environment of the terminal's location. This module integrates a light intensity sensor, anemometer, barometer, temperature and humidity sensor, and satellite imagery interface to construct a data environment parameter set centered on the current moment, encompassing both historical and real-time information. Feature parsing logic extracts core variables directly related to the intensity of new energy power generation from multi-source environmental parameters: light intensity data and wind speed data. Light intensity data includes historical hourly light intensity sequences for the past seven days, overlaid with cloud cover, relative humidity, and seasonal factors to form a multi-factor time series input matrix. Wind speed data includes historical wind speed curves, sudden wind event coefficients, and altitude-corrected atmospheric density variables. During data preprocessing, the light intensity sequence is standardized and imputed, then input into a time-series modeling unit based on a bidirectional long short-term memory (LSTM) network. The LSTM network uses a bidirectional LSTM structure to capture the forward evolution and backward backtracking features of the light intensity data in the time dimension, while also introducing attention weights after the hidden states. The recalculation mechanism assigns differentiated significance weights to different time segments, enabling the model to focus on seasonal, intermittent, and diurnal variation-significant periods, and output solar power predictions with enhanced time-series correlation. For wind speed data processing, historical wind speed values ​​are statistically summarized over multiple time periods to construct a probability distribution function suitable for wind resource characteristics. The frequency and abrupt changes of wind speed intervals are characterized by the fitting parameters of the distribution model. Subsequently, combined with the nonlinear output characteristics of wind power generation equipment, the wind speed intervals are mapped to the actual available wind power intervals. The mapping process considers the cubic relationship between wind speed and power and integrates physical boundary conditions such as wind turbine start-up and shutdown intervals and efficiency saturation values. The mapping between wind energy input and power output is realized through nonlinear curves, thereby forming predictable wind power output data per unit time. The project integrates the predicted renewable energy generation power with the remaining power data obtained from the current battery status module. An hourly time-series energy balance function is established, sequentially superimposing solar and wind power generation power, and subtracting the predicted power consumption from the system's internal load to construct an energy difference curve. Based on this, energy deficit identification and risk zone marking are performed. If the total input in a continuous time window is less than the predicted consumption, it is marked as a deficit segment; if the predicted renewable energy output consistently exceeds the load, it is marked as a surplus segment. The net energy change value within each window period is also statistically analyzed to construct energy balance data for the future time axis. The output energy balance prediction data includes the predicted power generation trend, load change prediction, remaining power support capacity analysis, and energy surplus / deficit zone division results.

[0029] In this embodiment, temporal feature extraction and attention weight calculation are performed on the illumination data to obtain solar power prediction results. This includes: constructing a dual-branch expert fusion memory network based on the heterogeneous features of illumination intensity changes under different weather patterns. The sunny day illumination expert branch uses an independent high-intensity illumination parameter model to extract direct light features, while the cloudy / rainy day illumination expert branch uses an independent low-intensity illumination parameter model to extract scattered light features. An expert gating mechanism is used to dynamically select and activate branches, resulting in a heterogeneous illumination expert network. Illumination data is then classified according to cloud cover rate and illumination intensity threshold, and input into the corresponding expert branches in the heterogeneous illumination expert network for deep feature extraction via convolutional neural networks and temporal modeling via long short-term memory networks, yielding... A multi-dimensional illumination temporal feature vector is generated. Position encoding and temporal embedding transformation are performed on this vector, and it is stored in a hierarchical illumination memory module via a hash indexing mechanism. Simultaneously, a mapping relationship between illumination patterns and historical power output is established to obtain a structured illumination historical feature memory. Based on this structured illumination historical feature memory and current environmental parameters including temperature, humidity, and atmospheric transparency, a multi-head attention mechanism is used to perform cross-temporal correlation fusion calculations. Residual connections and layer normalization are employed for feature enhancement to obtain fused illumination prediction features. These fused illumination prediction features are then input into an adaptive attention weight calculation module to identify key moment features. A multilayer perceptron decoder is used to perform nonlinear mapping and power regression calculations to obtain the solar power prediction results.

[0030] In one specific embodiment, the process of performing probability distribution modeling and nonlinear power curve mapping on wind speed data to obtain wind power prediction results can specifically include the following steps: Wind speed, gust coefficient, atmospheric pressure and air density are extracted from wind speed data and used to construct a wind energy prediction input vector. A wind turbine power curve model is established based on the relationship between air density and wind speed in the wind energy prediction input vector. Gradient vanishing compensation is performed on the wind energy prediction input vector to obtain the wind speed change feature vector; An attention mechanism is applied to the wind speed change feature vector to identify the moment of sudden wind speed change, and power calculation is performed by combining it with the wind turbine power curve model to obtain the wind power prediction result.

[0031] Specifically, the system jointly collects wind speed data for the terminal's location within a given period using a meteorological sensor network and a historical environmental database. This wind speed data includes real-time wind speed at ground level, historical wind speed variation sequences, gust index, atmospheric pressure corrected for geographical altitude, and current air temperature and humidity parameters. The corresponding air density value is then calculated, and the wind speed, gust coefficient, atmospheric pressure, and air density are aligned with a unified timestamp to construct a wind energy prediction input vector. The system accesses a wind turbine physical model database and selects a matching power curve model based on design parameters such as generator type, blade structure, rotor diameter, and rated output. The power curve maps air density and wind speed as primary variables to the output power space. During power curve fitting, the system considers the nonlinear variation characteristics of wind speed and applies hard constraints to boundary conditions such as turbine start-up wind speed, rated wind speed, and cut-out wind speed to ensure the model output has engineering-physical consistency. To address the vanishing gradient problem in deep network processing of wind speed time series data, a residual connection structure is introduced at the network front end. A gradient vanishing compensation mechanism is executed before the wind speed data enters the recursive neural unit. This mechanism maintains stable propagation of the wind speed change rate over long sequences through short-range connections and gradient adjustment within a sliding window, effectively preserving the continuity of gradual and abrupt wind speed changes. Based on this, a wind speed abrupt change detection channel is constructed from the processed wind speed change feature vector. An attention mechanism is used to calculate the importance weight of wind speed abrupt changes at each time step. Attention scores are calculated by combining local acceleration, slope, and variance in the wind speed sequence, thus identifying key moment segments of wind speed abrupt changes. The wind speed values ​​of these abrupt change segments are weighted and summarized to improve the model's sensitivity to extreme wind conditions. The attention-weighted wind speed change feature vector is then combined with a wind turbine power curve model for hourly power calculation. The wind speed input at each time step is fed into the corresponding segment of the power curve for mapping operations, outputting the predicted wind power value for the current time step. After predictions are completed at multiple time steps, the prediction results are reconstructed into a continuous wind power output time series.

[0032] In one specific embodiment, the process of performing step 102 may specifically include the following steps: Based on the energy budget forecast data, the collected data are prioritized to obtain emergency alarm data, abnormal event data, trend monitoring data, periodic routine data, and redundant background data. The emergency alarm data is subjected to lossless compression with CRC32 checksum to obtain the first compressed data; the abnormal event data is subjected to adaptive arithmetic coding to obtain the second compressed data; the trend monitoring data is subjected to piecewise linear fitting to obtain the third compressed data; the periodic routine data is subjected to incremental coding to obtain the fourth compressed data; and the redundant background data is subjected to spatiotemporal aggregation processing to obtain the fifth compressed data. The first compressed data, the second compressed data, the third compressed data, the fourth compressed data, and the fifth compressed data are encapsulated to obtain the identified compressed data; The compression level of the labeled compressed data is adjusted based on the energy budget forecast data to obtain the target compressed data.

[0033] Specifically, the multi-hour dynamic energy curve output by the energy balance prediction module is used as the basis for judgment. The energy surplus and deficit situation in each time window is analyzed. Different compression strategy intensity level thresholds are matched according to whether the energy balance is in the surplus, balance or deficit state in the next few hours. This threshold is used as a dynamic parameter to divide the priority of collected data. Through content recognition and feature matching mechanisms, the collected data is subjected to structured analysis to extract indicators such as data type, key field change rate, continuous time slice trend characteristics, and whether it has anomaly trigger tags. Based on these indicators, the timeliness, suddenness, and tolerability of the data are judged, and the data is divided into five levels: emergency alarm data, abnormal event data, trend monitoring data, periodic routine data, and redundant background data. Emergency alarm data includes equipment fault information, sensor over-limit alarms, and link interruption warnings. Abnormal event data includes sudden environmental parameters, boundary offsets, and unstable state records. Trend monitoring data mainly includes environmental status or equipment operating parameter curves collected continuously over multiple hours. Periodic routine data refers to repetitive data under stable conditions such as temperature, humidity, voltage, and current generated at regular intervals. Redundant background data refers to non-critical data with low information content, such as log information, repeated sampling, and debugging control bytes. For the data levels that have been classified, a compression strategy matching engine is connected. Different compression modules are called sequentially according to priority to compress the data content. Emergency alarm data enters the lossless compression module and embeds a CRC32 check field before compression to ensure that the data integrity is not affected. After lossless compression, the first compressed data is output. Abnormal event data enters the adaptive arithmetic coding module. The adaptive arithmetic coding module dynamically constructs an encoding table based on the symbol probability of the input data and performs entropy encoding to generate the second compressed data with a high compression ratio but stable decoding. Trend monitoring data is processed by a piecewise linear fitting processing engine to extract structural features, retaining only inflection points and key slope nodes to simplify the data structure and control fitting errors to generate the third compressed data. Periodic regular data is processed by the incremental coding module, which only calculates the changes between consecutive frames of data and performs difference encoding to generate the fourth compressed data. Redundant background data is input into the spatiotemporal aggregation processing module. By setting time windows and spatial variable intervals, several data samples are merged into statistical feature expressions such as mean, variance, and extreme values, and compressed into the fifth compressed data. All compressed data is uniformly input into the encapsulation module for structured identification. Each compressed data is appended with an identifier header consisting of a priority field, a timestamp field, a compression ratio field, and a verification flag. The mapping relationship between each type of compression method and the restoration function is recorded in the encapsulation table, forming recoverable identifiable compressed data.Read the latest energy balance forecast data, determine the compression intensity adjustment factor based on the energy status of the forecast interval. If the forecast is for an energy-sufficient interval, reduce the compression level of trend monitoring data and periodic routine data to retain more information details. If the forecast is for a near-deficit state, increase the compression level of periodic routine data and redundant background data to the maximum. At the same time, consider discarding non-critical fields or enabling super-aggregation mode to compress identifier fields to reduce volume. Under the action of the dynamic compression level control mechanism, output the target compressed data.

[0034] In one specific embodiment, the process of performing step 103 may specifically include the following steps: Extract data priority weight, queue delay time, and satellite transit time parameters from the target compressed data, and use the data priority weight, queue delay time, and satellite transit time parameters as input parameters; The weighting coefficients in the utility function are adjusted based on the remaining battery power data to obtain the battery power perception weighting coefficients. Input parameters and power perception weight coefficients are input into the utility function to calculate the utility value, thus obtaining the utility value of the transmission task; Task selection is performed based on the utility value of the transmission task to obtain a scheduling scheme.

[0035] Specifically, the structured encapsulation fields of data packets are read one by one from the target compressed data. The priority identifier field contained therein is extracted as the data priority weight. The weight value is assigned as a discrete value level according to the data type during the compression stage, and is distributed in a tiered manner according to the importance of the task. The cumulative waiting time of each data packet after entering the queue is read synchronously from the delay queue scheduling module as the queue delay time, and the time length unit is converted to minutes to unify the calculation input format. At the same time, through real-time analysis of satellite orbit data, the transit plans of all available satellites in the future period are obtained from the orbit prediction module driven by the dual-line element set. The corresponding satellite resources are matched according to the link node pointed to by the data packet scheduling target. The time difference between the current data packet and the start of the next available satellite window is calculated and marked as the satellite transit time parameter in minutes. The data priority weight, queue delay time and satellite transit time parameter are arranged in order and encapsulated as the input parameters of the utility function. The system obtains the current State of Charge (SOC) value and corresponding energy status identifier from the remaining power monitoring module. Combining this with the energy status range (sufficient, energy-saving, warning, or emergency), it calls the corresponding weight parameter correction factor from a preset weight coefficient adjustment table to dynamically adjust each control factor in the utility function. This constructs a power perception weight coefficient adapted to the current remaining power status. When power is sufficient, the emphasis on data integrity is increased while the weight coefficient for latency tolerance is reduced. Conversely, when power is in a warning or emergency state, protective scheduling of high-priority tasks is strengthened, and the threshold for time sensitivity is increased, thereby compressing the scheduling utility space of low-priority data. These input parameters and the power perception weight coefficient are then input into the utility function evaluation engine. The engine performs a non-linear mapping based on the task's current priority, waiting time in the queue, distance to the satellite window, and the current system's weight redistribution of these three factors. It outputs the utility value of the current task under the current energy constraints and communication timeliness conditions. This utility value is considered the priority for scheduling the task under the current conditions. After the utility values ​​of all tasks are calculated and sorted in batches, the task selection logic is executed by the scheduler. The scheduler evaluates all candidate tasks from high to low based on the upper limit of the energy budget for this schedulable transmission and the upcoming satellite window capacity limit, and prioritizes the task with the highest utility value to be packaged into the transmission candidate set. During the process of adding the candidate set, the scheduler dynamically checks whether the power consumption exceeds the budget and whether the transmission time matching constraint is met, and forms the optimal scheduling scheme under the current time slot.

[0036] In this embodiment, the weight coefficients in the utility function are adjusted based on the remaining power data to obtain the power perception weight coefficients. This includes: constructing a multi-state weight coefficient learning framework based on historical scheduling performance data under different remaining power states; extracting a shared weight adjustment representation function from historical scheduling data at different power levels using a state-invariant learning algorithm; and representing the optimal weight parameters under a specific power state as a combination of the shared function and power-related linear coefficients to obtain an adaptive weight learning model; inputting the current remaining power data and historical power consumption patterns into the adaptive weight learning model; eliminating the interference of different working environments on weight learning through a domain adversarial training mechanism; and extracting environment-independent weight adjustment strategy representations to obtain environment-independent weight features; and based on the environment-independent weight... A multi-objective weighted optimization problem of minimizing energy consumption and maximizing transmission efficiency is established based on key features. This problem is modeled as a coordinated control optimization among weight parameters α, β, and γ. The mutual constraints among the weight parameters are described by coupled state equations, resulting in a set of weighted coordinated control equations. An extended iterative algorithm is used to numerically solve the set of weighted coordinated control equations. By alternately updating the weight parameters and verifying the convergence of the system until a stable weight configuration equilibrium point is reached, a convergence proof is established to ensure the theoretical reliability of the algorithm, and the optimal weight parameter solution is obtained. Based on the optimal weight parameter solution and the current remaining percentage of power, the specific weight coefficients α, β, and γ in the utility function are calculated, and the weight coefficients are verified to meet the energy constraints and transmission performance requirements, resulting in the power perception weight coefficients.

[0037] In one specific embodiment, the process of selecting tasks based on the transmission task utility value to obtain a scheduling scheme may specifically include the following steps: Construct a utility sort queue by descending order of the utility values ​​of the transmission tasks; Calculate the transmission energy quota based on the remaining power data and the predicted average available power; Select transmission tasks one by one from the utility sorting queue and accumulate the transmission power consumption until the transmission energy quota is reached to obtain the transmission task set; The utility values ​​of the transmission tasks in the transmission task set are recalculated and queue adjustments are performed to obtain the scheduling scheme.

[0038] Specifically, the utility value of each transmission task is calculated one by one using a utility function corrected by the power sensing weight coefficient. The utility values ​​are then sorted from highest to lowest to construct a transmission task utility ranking queue, employing a stability-first descending order strategy. Remaining power data from the energy sensing module is retrieved and combined with the predicted average available power data output by the energy budget prediction module. Based on the maximum usable window duration within the current communication cycle, the protocol-specific transmission power consumption model, and the power safety threshold, the current available energy space and future replenishable energy are weighted and integrated to form the transmission energy quota for the current scheduling cycle. Tasks ranked highest are selected one by one from the ranked utility value queue, and their corresponding power consumption indicators are added to the current cumulative transmission power consumption. After each task is added, it is immediately checked whether the cumulative power consumption exceeds the current transmission energy quota. If it does not exceed the quota, the selection continues downwards; if it is about to exceed it, the selection stops, and the tasks currently included in the task set are marked as valid scheduling candidates, forming a dataset containing multiple high-value tasks within the energy budget—the transmission task set. The utility function calculation process is performed on each task in the transmission task set. The recalculation adjusts the input parameters based on real-time variables such as the latest energy state, latency growth, and window time decay. A difference analysis is performed between the newly calculated utility value and the previous utility value. When the change in utility value exceeds a preset threshold, a task reordering mechanism is triggered. Local fine-tuning or overall swapping is performed on the original order, and the reordering result is used as the structure output for the next scheduling list. After task reordering is complete, a scheduling scheme generation operation is performed. The updated task scheduling order in the transmission task set is combined with external boundary conditions such as the upcoming satellite window period, link transmission bandwidth, and single packet size limits to form the final scheduling plan. This plan includes parameters such as task ID, scheduling order, estimated power consumption, corresponding window start and end times, and the communication standard used.

[0039] In this embodiment, transmission tasks are selected one by one from the utility ranking queue, and the transmission power consumption is accumulated until the transmission energy quota is reached, resulting in a set of transmission tasks. This includes: a task selection optimization framework based on transmission service quality awareness; establishing quality constraints based on the minimum transmission success rate requirements of different priority transmission tasks in the utility ranking queue; setting a 99% success rate constraint for emergency alarm data, a 95% success rate constraint for abnormal event data, and a 90% success rate constraint for trend monitoring data, resulting in a hierarchical quality constraint system; calculating the expected transmission power consumption and success rate of each transmission task in the utility ranking queue under the current satellite signal strength and communication window conditions; and establishing a constraint optimization objective function that maximizes the accumulated transmission data volume under the hierarchical quality constraint system, resulting in quality-aware tasks. The selection model is chosen by converting the success rate constraint in the quality-aware task selection model into a penalty term using the Lagrange multiplier method, and then processing the nonlinear accumulation problem of transmission power consumption through logarithmic transformation to obtain a linearized selection optimization problem. Within the block coordinate descent iterative framework, the linearized selection optimization problem is decomposed into a transmission power consumption allocation subproblem and a task priority adjustment subproblem. A convex optimization algorithm is used to solve the power budget allocation and task sequence rearrangement respectively, and the selection parameters are iteratively updated until the transmission task set converges stably, yielding the optimal task selection strategy. Based on the optimal task selection strategy, transmission tasks satisfying the quality constraints are selected sequentially from the utility sorting queue according to the transmission power consumption accumulation order. Selection stops when the accumulated transmission power consumption approaches the 95% safety threshold of the transmission energy quota, resulting in the transmission task set.

[0040] In one specific embodiment, the process of performing step 104 may specifically include the following steps: Construct a delay queue based on the remaining battery power data and set a dynamic timeout threshold for the delay queue; Based on the scheduling scheme, the target compressed data is placed into the delay queue according to priority, and the satellite transit time and elevation angle parameters are calculated through the double-line element set data to obtain the effective communication window; Before the effective communication window begins, select the transmission task in the delay queue to obtain the transmission queue; Based on the remaining battery power data, the transmission mode is selected and the target compressed data in the transmission queue is sent to obtain the satellite transmission result.

[0041] Specifically, the remaining battery percentage reported by the current communication terminal is read, and a delay queue structure is initialized in the scheduling management module accordingly. This delay queue structure contains multiple independent channels, each corresponding to a data storage path with different priority levels. During construction, the timeout period for each channel is dynamically set based on the remaining battery value. When the remaining battery is high, low-priority data is allowed to have a longer waiting time. When the remaining battery is at a low to medium level or close to the critical threshold, the timeout period of the low-priority queue is compressed, ensuring that limited energy resources are concentrated on scheduling high-priority tasks. This achieves adaptive response to changes in battery status and optimal allocation of resources. Based on the priority information marked on each target compressed data in the scheduling scheme, the data is put into the corresponding priority channel according to predetermined logic, and the enqueue timestamp of each data item is recorded. The orbit analysis module parses the received dual-line element set orbit data and calculates all satellite transit periods that are relatively visible to the communication terminal within the next 24 hours based on the satellite dynamics model. By analyzing the real-time position changes of each satellite and the evolution trend of the ground elevation angle, the start and end times, peak elevation angle, and communication duration of each transit process are calculated. Combined with the preset available communication threshold and minimum elevation angle judgment criteria, an effective communication window table is constructed, recording all window segments that meet the communication conditions and their corresponding transmission capacity evaluation indicators. Before each upcoming communication window, the system enters a window preparation state, scans all pending data in the delay queue, prioritizes tasks in high-priority channels, and fills the transmission queue item by item using a utility value-based sorting strategy based on the remaining time and estimated transmission capacity of the current window. After each item is added, the system checks in real time whether the cumulative transmission time and energy consumption exceed the current window's capacity limit. If the limit is reached, task filling is terminated. After the transmission queue is built, the system calls the power management module again to obtain the latest remaining power value. Combining the current communication window's quality level, the physical characteristics of the satellite link, and the estimated total energy consumption of the task set, the system selects the most suitable transmission mode for the current conditions. When the energy status is good and the channel quality is excellent, the system prioritizes the higher throughput communication mode for batch data transmission. When the power is close to the lower limit or the window time slice is short, the system automatically switches to a low-power short message communication mode, completing the low-energy, high-efficiency transmission of tasks through a highly compressed data structure and a simplified protocol control process. During the mission transmission process, the transmission status is continuously monitored, and the status of each compressed data transmission is recorded to determine whether it was successful, whether a link interruption occurred, or whether a timeout occurred. After the transmission is completed, a satellite transmission result log is generated, which includes fields such as mission identifier, queue source, transmission mode, actual time taken, power consumption record, and result status.

[0042] In one specific embodiment, the process of selecting a transmission mode based on remaining battery power data and sending the target compressed data in the transmission queue to obtain the satellite transmission result can specifically include the following steps: Based on the remaining power data and energy harvesting power, determine whether to activate the shadow transmission mode and obtain the mode switching command; According to the transmission mode corresponding to the mode switching command, the trend monitoring data, periodic routine data and redundant background data in the transmission queue are converted into BeiDou short message format to obtain low-power transmission data. The transmission power consumption of low-power transmission data and conventional satellite communication data are calculated separately, and the optimal transmission path is selected to obtain the transmission path scheme. Data transmission is performed by implementing a transmission path scheme and monitoring the transmission success rate to obtain satellite transmission results.

[0043] Specifically, at the beginning of each scheduling cycle, the power sensing and energy assessment module is invoked to obtain the current remaining power percentage of the communication terminal and the statistical average of the new energy power collected over the past several consecutive hours. By setting judgment rules, these two input variables are jointly analyzed. If the remaining power is lower than the preset lower limit and the continuous energy input power is lower than the collection threshold, it is considered that the terminal is in a period of continuous energy deficit, thereby triggering the activation conditions of the shadow transmission mode. At the same time, a mode switching instruction containing the mode switching type, execution priority and working mode ID is output. The mode switching instruction is received by the scheduling management module and propagated in the entire data scheduling channel. After the system enters shadow mode, the scheduler filters the data content currently in the delay queue by type, identifying data packets belonging to trend monitoring, periodic routine, and redundant background categories. These three types of data are then input into the low-power format conversion unit as low-priority, high-redundancy data streams. In the low-power format conversion unit, the raw data is compressed and encapsulated according to the maximum transmission payload standard supported by the BeiDou short message communication protocol. Fixed-point format and differential expression are used to compress and map the data values, mapping each data to a compressed packet with a maximum size of no more than 78 bytes. At the same time, a dedicated message header structure is constructed, including device ID, collection timestamp, data type code, and integrity verification field, forming low-power transmission data. Low-power data and high-priority data retained in the conventional format are separately fed into the transmission power consumption modeling module. This module queries the protocol power consumption database based on the selected communication method (e.g., BeiDou, Tiantong, Iridium), and calculates the actual power consumption budget for both types of data under different link modes, considering parameters such as current RSSI signal strength, transmit power level, transmission duration, and data size. The power consumption results are then compared with the remaining energy quota. Simultaneously, the establishment delay, link availability, and retransmission success rate of each transmission path are considered. The path combination with the lowest power consumption and meeting the time window requirements is selected from all feasible paths to construct the transmission path scheme. The priority of the BeiDou short message system is increased to strengthen the low-power data delivery guarantee mechanism. After the path scheme is generated, data transmission operations are executed sequentially according to the path scheduling plan. During transmission, the channel status, response delay, ACK return, and retransmission behavior of each data item are monitored and recorded in real time. The transmission status is statistically analyzed, and a satellite transmission result table containing the transmitted data number, transmission path identifier, actual power consumption value, number of retransmissions, and final status is output.

[0044] The above describes the low-power satellite data acquisition terminal method powered by new energy sources in the embodiments of the present invention. The following describes the low-power satellite data acquisition terminal system powered by new energy sources in the embodiments of the present invention. Please refer to [link to relevant documentation]. Figure 2 One embodiment of the low-power satellite data acquisition terminal system powered by new energy sources in this invention includes: The monitoring module 201 is used to monitor the remaining power data of the communication terminal and calculate the energy income and expenditure prediction data in combination with environmental parameters; Compression module 202 is used to compress the collected data based on energy budget forecast data to obtain target compressed data; Create module 203 to create a scheduling scheme for the target compressed data; The transmission module 204 is used to allocate the target compressed data to the delay queue and match the valid communication window according to the scheduling scheme, obtain the transmission queue, select the transmission mode to send the target compressed data in the transmission queue, and obtain the satellite transmission result.

[0045] Through the collaborative efforts of the aforementioned components, combined with temperature resistance compensation and multi-sensor fusion, high-precision monitoring of remaining power is achieved, overcoming the large errors of traditional voltage detection methods. Predictive power output from solar and wind power is generated, automatically learning the periodic patterns of light intensity and the random characteristics of wind speed. A five-level data priority classification system is established, and specialized compression algorithms are used for data of different importance, maximizing transmission efficiency while ensuring the quality of critical data and avoiding resource waste caused by traditional uniform compression methods. Based on the energy marginal utility function, the value of data transmission is accurately quantified, comprehensively considering the dynamic balance between data importance, transmission delay, and energy consumption, overcoming the limitations of fixed scheduling strategies. Through dual-line element set orbit prediction and multi-level delay-tolerant queue management, accurate calculation of satellite transit times and full utilization of communication windows are achieved, solving the energy waste problem caused by blind transmission. In extremely low power conditions, the BeiDou short message backup transmission mechanism is activated, significantly reducing power consumption while ensuring reliable transmission of critical data. A refined energy allocation strategy and global optimization control mechanism are established, automatically selecting the optimal combination of working modes based on power status, maximizing communication time under energy constraints.

[0046] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0047] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part 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 the present invention. 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-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for a low-power satellite data acquisition terminal powered by a new energy source, characterized in that, include: Monitor the remaining battery power of communication terminals and calculate energy budget forecast data in combination with environmental parameters; Based on the energy budget prediction data, the collected data is compressed to obtain the target compressed data; Create a scheduling scheme for the target compressed data; According to the scheduling scheme, the target compressed data is allocated to the delay queue and matched with a valid communication window to obtain a transmission queue. The transmission mode is selected to send the target compressed data in the transmission queue, and the satellite transmission result is obtained.

2. The low-power satellite data acquisition terminal method for new energy power supply according to claim 1, characterized in that, The remaining battery power data of the monitoring communication terminal, combined with environmental parameters, is used to calculate energy budget prediction data, including: Acquire instantaneous current and temperature data of the battery in the communication terminal; Based on the battery temperature data and the number of charge-discharge cycles, the battery internal resistance is compensated and corrected to obtain temperature-corrected SOC data. A battery health state matrix is ​​constructed based on the instantaneous current data, the battery temperature data, and the temperature-corrected SOC data; The remaining available energy is calculated based on the battery health status matrix to obtain the remaining power data; Based on the remaining power data and environmental parameters, the power supply of new energy sources is predicted to obtain energy budget prediction data.

3. The low-power satellite data acquisition terminal method for new energy power supply according to claim 2, characterized in that, The step of predicting the power supply of new energy sources based on the remaining power data and environmental parameters to obtain energy budget prediction data includes: The environmental parameters of the communication terminal are obtained, and light and wind speed data are extracted from the environmental parameters. Temporal feature extraction and attention weight calculation are performed on the illumination data to obtain solar power prediction results; The wind speed data is modeled by probability distribution and mapped by nonlinear power curve to obtain wind power prediction results; By combining the remaining power data, energy balance calculations are performed on the solar power prediction results and the wind power prediction results to obtain energy budget prediction data.

4. The low-power satellite data acquisition terminal method powered by new energy sources according to claim 3, characterized in that, The process of performing probability distribution modeling and nonlinear power curve mapping on the wind speed data to obtain wind power prediction results includes: Wind speed, gust coefficient, atmospheric pressure, and air density are extracted from the wind speed data, and a wind energy prediction input vector is constructed. A wind turbine power curve model is established based on the relationship between air density and wind speed in the wind energy prediction input vector. Gradient vanishing compensation is performed on the wind energy prediction input vector to obtain the wind speed change feature vector; An attention mechanism is applied to the wind speed change feature vector to identify the moment of sudden wind speed change, and power is calculated by combining it with the wind turbine power curve model to obtain the wind power prediction result.

5. The low-power satellite data acquisition terminal method for new energy power supply according to claim 1, characterized in that, The step of compressing the collected data based on the energy budget forecast data to obtain the target compressed data includes: Based on the energy budget forecast data, the collected data is prioritized to obtain emergency alarm data, abnormal event data, trend monitoring data, periodic routine data, and redundant background data. The emergency alarm data is subjected to lossless compression with CRC32 checksum to obtain the first compressed data; the abnormal event data is subjected to adaptive arithmetic coding to obtain the second compressed data; the trend monitoring data is subjected to piecewise linear fitting to obtain the third compressed data; the periodic regular data is subjected to incremental coding to obtain the fourth compressed data; and the redundant background data is subjected to spatiotemporal aggregation processing to obtain the fifth compressed data. The first compressed data, the second compressed data, the third compressed data, the fourth compressed data, and the fifth compressed data are respectively encapsulated to obtain the identified compressed data; The compression level of the labeled compressed data is adjusted based on the energy budget forecast data to obtain the target compressed data.

6. The low-power satellite data acquisition terminal method for new energy power supply according to claim 1, characterized in that, The scheduling scheme for creating the target compressed data includes: Extract data priority weight, queue delay time, and satellite transit time parameters from the target compressed data, and use the data priority weight, queue delay time, and satellite transit time parameters as input parameters; The weighting coefficients in the utility function are adjusted based on the remaining power data to obtain the power perception weighting coefficients. The input parameters and the power perception weight coefficient are input into the utility function to calculate the utility value, and the utility value of the transmission task is obtained. Based on the utility value of the transmission task, a task selection is performed to obtain a scheduling scheme.

7. The low-power satellite data acquisition terminal method for new energy power supply according to claim 6, characterized in that, The step of selecting tasks based on the transmission task utility value to obtain a scheduling scheme includes: Construct a utility sorting queue by sorting the utility values ​​of the transmission tasks in descending order; The transmission energy quota is calculated based on the remaining power data and the predicted average available power. The transmission tasks are selected one by one from the utility sorting queue and the transmission power consumption is accumulated until the transmission energy quota is reached, thus obtaining a set of transmission tasks. The utility values ​​of the transmission tasks in the transmission task set are recalculated and queue adjustments are performed to obtain a scheduling scheme.

8. The low-power satellite data acquisition terminal method for new energy power supply according to claim 1, characterized in that, The process of allocating the target compressed data to a delay queue and matching it with a valid communication window according to the scheduling scheme, obtaining a transmission queue, selecting a transmission mode to send the target compressed data in the transmission queue, and obtaining the satellite transmission result includes: A delay queue is constructed based on the remaining battery power data, and a dynamic timeout threshold is set for the delay queue; Based on the scheduling scheme, the target compressed data is placed into the delay queue according to priority, and the satellite transit time and elevation angle parameters are calculated through the dual-line element set data to obtain the effective communication window; Before the start of the effective communication window, select the transmission task in the delay queue to obtain the transmission queue; Based on the remaining battery power data, a transmission mode is selected and the target compressed data in the transmission queue is sent to obtain the satellite transmission result.

9. The low-power satellite data acquisition terminal method for new energy power supply according to claim 8, characterized in that, The process of selecting a transmission mode based on the remaining battery power data and sending the target compressed data in the transmission queue to obtain the satellite transmission result includes: Based on the remaining power data and energy harvesting power, determine whether to activate the shadow transmission mode and obtain a mode switching command; According to the transmission mode corresponding to the mode switching instruction, the trend monitoring data, periodic regular data and redundant background data in the transmission queue are converted into BeiDou short message format to obtain low-power transmission data. The transmission power consumption of the low-power transmission data and the conventional satellite communication data are calculated respectively, and the optimal transmission path is selected to obtain the transmission path scheme; Data transmission is performed using the aforementioned transmission path scheme, and the transmission success rate is monitored to obtain the satellite transmission results.

10. A low-power satellite data acquisition terminal system powered by a new energy source, characterized in that, A method for executing a low-power satellite data acquisition terminal powered by a new energy source as described in any one of claims 1-7, comprising: The monitoring module is used to monitor the remaining battery power of the communication terminal and calculate the energy budget forecast data in combination with environmental parameters. The compression module is used to compress the collected data based on the energy budget prediction data to obtain the target compressed data; A creation module is used to create a scheduling scheme for the target compressed data; The transmission module is used to allocate the target compressed data to a delay queue and match a valid communication window according to the scheduling scheme, obtain a transmission queue, select a transmission mode to send the target compressed data in the transmission queue, and obtain the satellite transmission result.

Images