A satellite-ground integrated Internet of Things communication processing method and system based on multi-network integration

Abstract

The present invention relates to the field of satellite Internet of Things technology, and in particular to a satellite-ground integrated Internet of Things communication processing method and system based on multi-network integration, the method specifically comprising: obtaining the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite, switching to a backup network mode, the backup network mode including a cellular network, a LoRa network and a 6G network, and simultaneously identifying the fault type label of the first data to obtain the second data; constructing an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of the energy required for data regeneration; and determining the fault self-healing scheme of the satellite-ground integrated Internet of Things terminal by constructing an energy scheduling model, the fault type and the fault location through the multi-layer perceptron neural network. The present invention improves the adaptive ability and self-healing ability of the satellite-ground integrated Internet of Things system when encountering data transmission abnormalities, and optimizes energy management.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite Internet of Things, and in particular to a satellite-ground integrated Internet of Things communication processing method and system based on multi-network integration. Background Art

[0002] With the rapid development of Internet of Things technology, satellite-ground integrated Internet of Things communication, as one of the core directions of future communication technology, is gradually becoming a key technology to achieve global seamless connection and efficient data transmission. Traditional Internet of Things communication methods mainly rely on ground networks, such as cellular networks, LoRa networks, etc., but in remote areas, oceans and high altitude environments, the coverage and communication quality of ground networks are severely limited. Therefore, the satellite-ground integrated Internet of Things communication mode combined with low-orbit satellites came into being, aiming to make up for the shortcomings of the ground network through satellite networks and realize global information interconnection.

[0003] However, satellite-to-ground IoT communications still face many challenges in practical applications. First, there are significant differences between satellite networks and ground networks in terms of protocols, frequencies, and transmission rates. How to achieve multi-network integration and ensure the continuity and stability of data transmission has become an urgent problem to be solved. Secondly, satellite-to-ground communication links are easily affected by factors such as atmospheric conditions and changes in satellite orbits, resulting in unstable communication link quality and possible abnormalities during data transmission. In addition, the energy supply of IoT terminal devices is limited. How to achieve efficient energy utilization and fault self-healing while ensuring communication quality is also a technical problem that needs to be solved urgently in satellite-to-ground IoT communications. Summary of the invention

[0004] The purpose of the present invention is to provide a satellite-ground integrated Internet of Things communication processing method based on multi-network integration, which significantly improves the adaptability and self-healing ability of the satellite-ground integrated Internet of Things system when encountering data transmission anomalies, optimizes energy management, reduces energy consumption, and improves the operating efficiency and reliability of the system, so as to solve at least one of the above-mentioned prior art problems.

[0005] In a first aspect, a satellite-ground integrated Internet of Things communication processing method based on multi-network integration is provided, and the method specifically comprises:

[0006] Acquire the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting first data to the low-orbit satellite, switch to a backup network mode, where the backup network mode includes a cellular network, a LoRa network, and a 6G network, and perform fault type label identification on the first data to obtain second data;

[0007] Acquire environmental data between a ground control center and a low-orbit satellite, determine a predicted value of energy required for data regeneration of the second data after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determine a data regeneration processing scheme for data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration;

[0008] Building an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration;

[0009] Regenerate the second data according to the data regeneration processing scheme to obtain a data regeneration result, and based on the data regeneration result, use a support vector machine fault diagnosis algorithm to identify the fault type and fault location of the satellite-ground integrated Internet of Things terminal;

[0010] Calculate the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determine the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location;

[0011] The energy supply conditions of each satellite-ground integrated Internet of Things terminal are obtained, and according to the fault self-healing scheme and the energy supply conditions, a dynamic programming algorithm is used to dynamically adjust the energy allocation scheme between each satellite-ground integrated Internet of Things terminal.

[0012] In a second aspect, a satellite-ground integrated Internet of Things communication processing system based on multi-network integration is provided, wherein the system specifically comprises:

[0013] A first processing module is used to obtain the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite, switch to a backup network mode, where the backup network mode includes a cellular network, a LoRa network, and a 6G network, and at the same time, perform fault type label identification on the first data to obtain second data;

[0014] a second processing module, configured to obtain environmental data between the ground control center and the low-orbit satellite, determine a predicted value of energy required for data regeneration of the second data after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determine a data regeneration processing scheme for data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration;

[0015] A third processing module is used to construct an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration;

[0016] a fourth processing module, configured to perform data regeneration on the second data according to the data regeneration processing scheme to obtain a data regeneration result, and identify a fault type and a fault location of the satellite-ground integrated Internet of Things terminal by using a support vector machine fault diagnosis algorithm based on the data regeneration result;

[0017] A fifth processing module is used to calculate the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determine the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location;

[0018] The sixth processing module is used to obtain the energy supply status of each satellite-ground integrated Internet of Things terminal, and dynamically adjust the energy allocation plan between each satellite-ground integrated Internet of Things terminal according to the fault self-healing plan and the energy supply status by using a dynamic programming algorithm.

[0019] In a third aspect, the present invention provides a computer device comprising: a memory and a processor and a computer program stored in the memory. When the computer program is executed on the processor, it implements a satellite-ground integrated Internet of Things communication processing method based on multi-network integration as described in any one of the above methods.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program implements a satellite-ground integrated Internet of Things communication processing method based on multi-network integration as described in any one of the above methods.

[0021] Compared with the prior art, the present invention has at least one of the following technical effects:

[0022] 1. Significantly improve the adaptive and self-healing capabilities of the satellite-ground integrated IoT system when encountering data transmission anomalies, optimize energy management, reduce energy consumption, and improve the system's operational efficiency and reliability.

[0023] 2. The present invention can effectively ensure the stable operation of the satellite-ground integrated Internet of Things terminal through intelligent fault handling and energy scheduling, which is of great significance for ensuring the integrity and timeliness of key data transmission.

[0024] 3. The present invention can obtain the location information of the satellite-ground integrated IoT terminal in real time, and quickly switch to the backup network mode (including cellular network, LoRa network and 6G network) when data transmission is abnormal, to ensure the continuity and stability of data transmission. This multi-network integration and seamless switching mechanism effectively improves the reliability and flexibility of communication.

[0025] 4. The present invention predicts the energy required for data regeneration by acquiring environmental data and formulates a data regeneration processing plan accordingly, thereby achieving efficient recovery of damaged data.

[0026] 5. The present invention combines a multi-layer perceptron neural network and a support vector machine fault diagnosis algorithm to accurately identify the fault type and fault location of the regenerated data, providing strong support for subsequent fault self-healing.

[0027] 6. The present invention calculates the energy consumption value required for fault self-healing based on the energy scheduling model, and formulates a fault self-healing plan in combination with the fault type and fault location. In combination with the energy supply situation of each terminal, a dynamic programming algorithm is used to dynamically adjust the energy allocation plan to achieve efficient energy utilization and intelligent management of fault self-healing, thereby improving the reliability and service life of the Internet of Things terminals and reducing the overall operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 It is a flowchart of a satellite-ground integrated Internet of Things communication processing method based on multi-network integration provided by an embodiment of the present invention;

[0030] Figure 2 It is a structural schematic diagram of a satellite-ground integrated Internet of Things communication processing system based on multi-network integration provided by an embodiment of the present invention;

[0031] Figure 3 It is a structural schematic diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0033] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0034] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0035] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0036] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0037] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0038] In the embodiment of the present application, the execution subject of the process includes a terminal device, which includes but is not limited to: a server, a computer, a smart phone, a tablet computer, and other devices capable of executing the method disclosed in the present application. Figure 1 A schematic flow chart of a satellite-ground integrated IoT communication processing method based on multi-network integration disclosed in an embodiment of the present invention is shown, and the details are as follows:

[0039] S101, obtaining the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs in the process of the satellite-ground integrated Internet of Things terminal transmitting first data to the low-orbit satellite, switching to a backup network mode, wherein the backup network mode includes a cellular network, a LoRa network, and a 6G network, and at the same time, performing a fault type label identification on the first data to obtain second data.

[0040] In this embodiment, the satellite-ground integrated Internet of Things terminal is equipped with a GPS module, a Beidou positioning module and a variety of communication modules (including a low-orbit satellite communication module, a cellular network module, a LoRa network module and a 6G network module). The satellite-ground integrated Internet of Things fusion system obtains the real-time location information of each terminal through the GPS module or the Beidou positioning module, and monitors its data transmission status.

[0041] The satellite-ground integrated IoT terminal realizes dual-network intelligent signal search and switching or dual-network dual-standby of ground LTE network and satellite IoT through cat1 IoT and Soc intelligent module frequency modulation baseband technology, and supports LTE IoT in areas with LTE base stations and UHF satellite IoT in areas without LTE base stations. While reducing power consumption through dual-network intelligent switching, it selects the network service with the best cost to realize global all-weather IoT data backhaul services.

[0042] The satellite-ground integrated IoT fusion system realizes multi-band intelligent algorithm recognition. The data platform automatically identifies the frequency band source and quickly responds to data classification source management. It has the characteristics of dual-network intelligent switching or dual-network dual-standby data backhaul backup network. It can fully empower LTE IoT products with power and network, and realize the transition from traditional ground-based LTE IoT to "LTE+satellite" satellite-ground integrated IoT.

[0043] While data is switching and transmitting, the satellite-ground integrated IoT fusion system identifies the fault type label of the original first data. Through the built-in fault diagnosis algorithm, the system determines that the data anomaly is due to communication interruption caused by signal interference, and marks the data with a corresponding fault label to generate the second data.

[0044] In the process of transmitting the first data from the satellite-ground integrated IoT terminal to the low-orbit satellite, the transmission status is monitored in real time, and the abnormality detection algorithm is used to determine whether an abnormality occurs. If an abnormality occurs, it is switched to the backup network mode, which includes cellular network, LoRa network and 6G network. After switching to the backup network mode, the first data is input into the fault type label recognition model, and the support vector machine algorithm is used to multi-classify the fault type to obtain the fault type label, and the fault type label is attached to the first data to form the second data. According to the fault type label, the optimal transmission scheme is determined by the decision tree algorithm, and one or more are selected from the cellular network, LoRa network and 6G network as the backup transmission channel. The second data is transmitted through the backup transmission channel, and the second data is encrypted by the data encryption algorithm to ensure the security of data transmission. After the transmission of the backup transmission channel is completed, the transmission quality is evaluated, and the transmission delay, packet loss rate and other indicators are analyzed by the clustering algorithm. The priority of the backup transmission channel is dynamically adjusted according to the analysis results to ensure the stability and reliability of the transmission. The second data is stored in a distributed database, and the second data is compressed and stored by the data compression algorithm to improve the storage efficiency. At the same time, an index of the second data is established, and the inverted index technology is used to speed up data retrieval. Finally, the second data is visualized, and a chart is generated through a data visualization algorithm to intuitively present the data transmission process and fault conditions, which is convenient for data analysis and decision optimization.

[0045] In this embodiment, the location information of the IoT terminal is obtained in real time through the GPS module or Beidou positioning, so that the system can quickly locate the device with communication anomalies, providing basic data support for subsequent fault handling and network switching. When data transmission is abnormal, the satellite-ground integrated IoT fusion system can intelligently switch to the backup network mode automatically according to the current network environment, ensuring continuous data transmission. This multi-network fusion and intelligent switching mechanism significantly improves the reliability and stability of communication.

[0046] For example, the satellite-ground integrated IoT terminal obtains real-time location information through the GPS module, including longitude, latitude, altitude and other data, in JSON format as the first data. In the process of transmitting to the low-orbit satellite, by monitoring the round-trip time and packet loss rate of the data packet, the isolation forest algorithm is used for anomaly detection. If the round-trip time exceeds 500ms or the packet loss rate is higher than 5%, it is judged as abnormal. At this time, switch to the backup network mode, use the random forest algorithm to multi-classify the fault type, identify the fault type such as signal attenuation, equipment failure or network congestion, generate fault type labels such as "signal_loss", etc., and attach them to the first data to form the second data. According to the fault type label, the CART decision tree algorithm is used to comprehensively consider factors such as transmission rate, power consumption, coverage, etc. to determine the optimal transmission plan, such as selecting a combination of 6G network and LoRa network. When transmitting the second data in the backup channel, the AES-256 encryption algorithm is used to ensure data security. After the transmission is completed, the K-means clustering algorithm is used to perform cluster analysis on indicators such as transmission delay and packet loss rate. If the delay is less than 100ms and the packet loss rate is less than 1%, it is a high-quality transmission, and the priority of the backup channel is increased; otherwise, the priority is lowered to dynamically optimize the transmission strategy. The second data is stored in the Apache Cassandra distributed database and compressed using the LZ4 compression algorithm with a compression ratio of 3:1. At the same time, Lucene is used to build an inverted index to shorten the data retrieval time to milliseconds. Finally, through data visualization libraries such as ECharts, a time sequence diagram of the transmission process and a pie chart of the fault type are generated to intuitively display the data transmission and fault conditions, providing data support for optimization decisions.

[0047] In some embodiments, when an abnormality occurs during the process of transmitting the first data from the satellite-ground integrated Internet of Things terminal to the low-orbit satellite, switching to the backup network mode specifically includes:

[0048] Determine the coordinate information of the satellite-ground integrated Internet of Things terminal according to the real-time position, insert the coordinate information into the data packet header of the first data, and transmit it to the ground control center;

[0049] During the transmission of the first data, network status parameters are collected in real time, and according to the network status parameters, a hidden Markov model is used to model the network status between the satellite-ground integrated Internet of Things terminal and the low-orbit satellite to obtain a network status hidden Markov model;

[0050] The Viterbi algorithm is used to solve the network state hidden Markov model to obtain the optimal state transfer path of the network state. According to the optimal state transfer path of the network state, it is determined whether an abnormality occurs in the process of the current satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite. If it is determined to be abnormal, the backup network mode is used to continue the transmission of the first data.

[0051] In this embodiment, according to the real-time location information of the satellite-ground integrated Internet of Things terminal, the longitude and latitude coordinates of the terminal are obtained by satellite positioning technology. By converting the acquired longitude and latitude coordinates into a standard geographic coordinate format, the standardized coordinate information of the satellite-ground integrated Internet of Things terminal is obtained. According to the data transmission protocol specification, the standardized coordinate information of the satellite-ground integrated Internet of Things terminal is inserted into the data packet header of the first data. Using a satellite communication link, the first data with the coordinate information inserted is transmitted from the satellite-ground integrated Internet of Things terminal to the ground control center. The location information and original business data of the satellite-ground integrated Internet of Things terminal are obtained through the first data with the coordinate information inserted received by the ground control center. According to the received location information of the satellite-ground integrated Internet of Things terminal, the geographical location distribution of the terminal is determined. According to the content of the original business data and the location distribution of the satellite-ground integrated Internet of Things terminal, the spatiotemporal attributes of the business data are judged. By analyzing the spatiotemporal attributes of the business data, the spatiotemporal distribution characteristics of the satellite-ground integrated Internet of Things terminal business are obtained.

[0052] According to the network status between the satellite-ground integrated IoT terminal and the low-orbit satellite, network status parameters are collected in real time during the transmission of the first data, including network delay, packet loss rate, bandwidth and other indicators, to obtain an observation sequence reflecting the network status. By preprocessing and feature extraction of the collected network status parameters, an observation sequence suitable for the hidden Markov model is obtained. According to the observation sequence, the parameters of the hidden Markov model are estimated using the Baum-Welch algorithm, including the initial state probability distribution, the state transition probability matrix and the observation probability matrix, and the optimal model parameters are obtained through iterative optimization. The estimated model parameters are substituted into the hidden Markov model to construct a hidden Markov model of the network status between the satellite-ground integrated IoT terminal and the low-orbit satellite, which can characterize the dynamic change characteristics of the network status. The network status is modeled using the network status hidden Markov model to obtain the state transition probability matrix and the observation probability matrix. By real-time monitoring the network status observation sequence in the process of the satellite-ground integrated IoT terminal transmitting the first data to the low-orbit satellite, the constructed network status hidden Markov model is solved using the Viterbi algorithm to obtain the optimal state transition path of the network status. According to the obtained optimal state transfer path of the network state, determine whether there is an abnormality in the process of the current satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite, and determine the reliability and stability of the network transmission. If the judgment result is that the current network transmission process is abnormal, the backup network mode is used to continue to transmit the first data to ensure the integrity and continuity of data transmission. The first data is transmitted through the backup network mode, and the completion status and quality assessment results of the data transmission are obtained to determine whether the transmission performance under the backup network mode meets the business requirements. According to the transmission of the first data in the main network mode and the backup network mode, the quality of the communication link between the satellite-ground integrated Internet of Things terminal and the low-orbit satellite is comprehensively evaluated to obtain the optimization adjustment strategy of the network state.

[0053] In some embodiments, performing fault type label identification on the first data to obtain second data specifically includes:

[0054] Obtain historical data transmitted by the satellite-ground integrated IoT terminal to the low-orbit satellite, add a fault type label to each abnormal data in the historical data, and form a fault diagnosis data set;

[0055] Extracting key features from the fault diagnosis data set to obtain feature vectors, training the feature vectors using a support vector machine algorithm to obtain a fault diagnosis feature space representation;

[0056] After associating the fault diagnosis feature space representation with each fault type, an associated data set is formed, and the associated data set is used as input to train a decision tree model to establish a fault type diagnosis model;

[0057] The fault type label of the first data is identified according to the fault type diagnosis model to obtain second data.

[0058] In this embodiment, according to the communication protocol between the satellite-ground integrated IoT terminal and the low-orbit satellite, the historical data transmitted by the IoT terminal to the low-orbit satellite is obtained to obtain the original IoT terminal operation data set. The data cleaning and preprocessing technology is used to perform denoising, normalization and other processing on the obtained original IoT terminal operation data to obtain a high-quality IoT terminal operation data set. Through expert knowledge and fault mode analysis, the common fault types of IoT terminals are obtained, and the discrimination rules and threshold conditions for each fault type are determined. A rule-based anomaly detection algorithm is used to match and compare each data in the IoT terminal operation data set with the fault discrimination rule to determine whether the data is abnormal data and obtain the abnormal label of the data. Through the abnormal data identification result, all abnormal data in the IoT terminal operation data set are obtained, and the specific fault type to which each abnormal data belongs is determined according to the fault discrimination rule. Data annotation technology is used to add the corresponding fault type label to each abnormal data to obtain an abnormal data set with a fault label. The abnormal data set with the fault label is merged with the original IoT terminal operation data set to obtain a complete IoT terminal operation data set with a fault label to form the final fault diagnosis data set.

[0059] According to the fault diagnosis business needs, data acquisition equipment is used to obtain raw data related to fault diagnosis. The raw data is cleaned and standardized through data preprocessing methods to obtain a high-quality fault diagnosis data set. According to the characteristics of the fault diagnosis data set, feature engineering methods are used to analyze the data set. Through feature selection and feature extraction techniques, key features that can characterize the fault characteristics are extracted from the fault diagnosis data set to obtain feature vectors with strong representativeness and high discrimination. Based on the obtained feature vectors, the support vector machine algorithm is used to train the feature vectors. By optimizing the parameters of the support vector machine model, a fault diagnosis feature space representation that can efficiently and accurately diagnose and classify faults is obtained.

[0060] According to the relationship between the fault diagnosis feature space representation and each fault type, the associated data set is obtained and used as the input data of the decision tree model. The decision tree algorithm is used to train the associated data set, and the decision tree is constructed by recursive partitioning. At each node, an optimal feature is selected as the partition attribute, and the data set is divided into several subsets according to the value of the attribute. The divided sub-datasets are obtained, and the above partitioning process is recursively repeated for each sub-dataset until the stop condition is met to obtain the leaf nodes of the decision tree. By marking the leaf nodes of the decision tree with categories, each leaf node is associated with the corresponding fault type to form a fault type diagnosis rule. The generated decision tree model is used to classify the new fault diagnosis feature space representation. By traversing the judgment nodes of the decision tree, branch selection is performed according to the value of the feature attribute, and finally the leaf node is reached and the associated fault type is obtained to achieve fault diagnosis. According to the diagnosis results, the fault type of the equipment is judged, and corresponding maintenance or treatment measures are taken to improve the reliability and availability of the equipment. By continuously accumulating fault diagnosis data, the decision tree model is updated and optimized to improve the accuracy and efficiency of fault diagnosis. An optimized fault type diagnosis model is obtained, which is used to identify the fault type label of the first data to obtain second data.

[0061] S102, acquiring environmental data between the ground control center and the low-orbit satellite, determining a predicted value of energy required for data regeneration of the second data after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determining a data regeneration processing scheme for data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration.

[0062] In this embodiment, the ground control center collects environmental data related to the low-orbit satellite communication link through the deployed sensor network, such as atmospheric conditions (such as cloud thickness, rainfall probability), electromagnetic interference level, satellite orbit parameters, etc., and these data are transmitted to the data processing system of the ground control center in real time. The data processing system uses advanced algorithm models to predict the second data (i.e., the data after fault type label identification) after being transmitted from the low-orbit satellite to the ground control center based on the collected environmental data, and considers various factors such as signal attenuation, bit error rate, data transmission rate, and the computing power of the ground control center to predict the energy required for data regeneration. Based on the predicted value of energy required for data regeneration, the data processing system further determines the processing scheme for data regeneration of the second data at the ground control center. This processing scheme includes selecting the optimal data regeneration algorithm, allocating sufficient computing resources, optimizing the data transmission protocol, etc., to ensure the highest data recovery rate and accuracy under limited energy consumption.

[0063] Specifically, the environmental data is modeled and analyzed by a support vector machine regression algorithm to obtain channel transmission characteristics between the low-orbit satellite and the ground control center, including parameters such as channel fading coefficient and signal-to-noise ratio. According to the channel transmission characteristics, the Monte Carlo simulation method is used to simulate the process of transmitting the second data from the low-orbit satellite to the ground control center, and the bit error rate curve during the second data transmission process is obtained. According to the bit error rate curve, the bit error rate prediction value after the second data is transmitted from the low-orbit satellite to the ground control center is obtained by least squares fitting. According to the Shannon information theory formula, by substituting the bit error rate prediction value into the formula, the data regeneration energy prediction value after the second data is transmitted from the low-orbit satellite to the ground control center is calculated. According to the data regeneration energy prediction value, the dynamic programming algorithm is used to optimize and solve the process of data regeneration of the second data in the ground control center, and the optimal data regeneration processing scheme for the second data is obtained, including parameters such as signal power, coding mode, and modulation mode during data regeneration. The optimal data regeneration processing scheme is applied to the ground control center to regenerate the second data, and the regenerated second data is obtained as the input of the subsequent task planning.

[0064] In this embodiment, by collecting and analyzing environmental data between the ground control center and the low-orbit satellite, the system can accurately predict the energy consumption required for data regeneration, which helps to make more reasonable decisions under limited resources and avoid unnecessary energy waste. The data regeneration processing solution formulated based on the energy prediction value can ensure the maximum utilization of energy while meeting the data recovery rate and accuracy, which not only improves the efficiency of data processing, but also reduces operating costs.

[0065] In some embodiments, determining the predicted value of energy required for data regeneration after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data specifically includes:

[0066] Acquire historical transmission loss degree data of second data transmitted from the low-orbit satellite to the ground control center, and determine a data set of energy value required for data regeneration according to the historical transmission loss degree data;

[0067] Associating the environmental data with the data set of energy value required for data regeneration, and inputting them into a support vector machine regression model or a long short-term memory neural network model for training to obtain a prediction model of energy required for data regeneration;

[0068] The data regeneration required energy prediction value after the current second data is transmitted from the low-orbit satellite to the ground control center is determined according to the data regeneration required energy prediction model.

[0069] In this embodiment, based on the second data transmitted from the low-orbit satellite to the ground control center, the historical transmission loss degree data of the data is obtained, and the historical transmission loss degree data reflects the signal attenuation and loss of the satellite data during the transmission process. The obtained historical transmission loss degree data is processed by a statistical analysis method to obtain statistical characteristics such as the mean and variance of the transmission loss degree, which are used to characterize the overall distribution of the transmission loss. By constructing a data regeneration energy consumption model, the historical transmission loss degree data is used as input, and a mapping relationship between the transmission loss degree and the energy required for data regeneration is established to obtain a prediction model for data regeneration energy consumption. According to the data regeneration energy consumption model, a numerical calculation method is used to calculate the corresponding data regeneration required energy value for different historical transmission loss degree data, and obtain a data regeneration required energy value data set.

[0070] By associating the preprocessed environmental data with the data set of energy values ​​required for data regeneration, training samples are constructed. Each sample contains environmental data as input features and the corresponding energy value required for data regeneration as the target output. According to business needs and data characteristics, a support vector machine regression model or a long short-term memory neural network model is selected as the basic algorithm of the data regeneration energy prediction model. The associated training samples are input into the selected model, and the model is trained by setting appropriate hyperparameters and training iterations to enable it to capture the complex relationship between environmental data and the energy value required for data regeneration. During the model training process, cross-validation and other techniques are used to evaluate the generalization performance of the model, and the model structure and hyperparameters are adjusted according to the evaluation results to obtain better prediction results. Through multiple iterations of training and tuning, an excellent performance data regeneration energy prediction model is obtained, which can accurately predict the energy value required in the data regeneration process based on the input environmental data. Through the data regeneration energy prediction model, the data regeneration energy prediction value after the second data collected by the current low-orbit satellite is transmitted to the ground control center is determined.

[0071] In some embodiments, determining the data regeneration processing scheme for performing data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration specifically includes:

[0072] Taking the predicted value of energy required for data regeneration as a constraint condition, and taking the quality requirement after data regeneration and the efficiency requirement during efficiency regeneration as an objective function;

[0073] According to the constraint conditions and the objective function, a multi-objective optimization algorithm is used to calculate and obtain a data regeneration processing plan for regenerating the second data at the ground control center.

[0074] In this embodiment, according to the predicted value of energy required for data regeneration as a constraint condition, the quality requirements after data regeneration and the efficiency requirements during efficiency regeneration are used as the objective function to obtain the input parameters of the data regeneration task. By analyzing the constraints and objective functions of the data regeneration task, according to the characteristics of the data regeneration task, a genetic algorithm is used as a specific multi-objective optimization algorithm, and the parameters in the data regeneration process are encoded as chromosomes. By designing a suitable fitness function, the requirements for data regeneration quality and efficiency are mapped to the fitness value of the chromosome, and a quantitative index for evaluating the quality of the chromosome is obtained. The chromosome population is iteratively optimized using genetic operators such as selection, crossover, and mutation to obtain an optimized solution for the data regeneration parameters, and the constraints on data regeneration energy consumption and the improvement of quality and efficiency are achieved. According to the optimized data regeneration parameters, a data regeneration system is used to perform the data regeneration task to obtain high-quality data after regeneration. By evaluating the quality indicators and regeneration efficiency of the regenerated data, it is determined whether the results of data regeneration meet the expected requirements. According to the feedback of the data regeneration results, the iterative optimization mechanism of the genetic algorithm is used to continuously adjust the data regeneration parameters to obtain a better regeneration solution and continuously improve the performance of data regeneration. Through multiple rounds of iterative optimization, the optimal data regeneration parameter combination that meets the energy consumption constraints and quality and efficiency requirements is obtained, and the final execution plan of the data regeneration task is determined.

[0075] S103: constructing an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration.

[0076] In this embodiment, a multi-layer perceptron (MLP) neural network is used to construct an energy scheduling model. The model uses the fault type label of the second data and the predicted value of the energy required for data regeneration as input features, and learns and simulates the optimal energy allocation strategy under different fault types and energy requirements through nonlinear transformations of multiple hidden layers. During the training process, historical data and simulation scenarios are used to train the MLP neural network, and the network parameters are continuously adjusted to optimize the performance of the energy scheduling model. After training, the model can quickly give energy scheduling suggestions based on new input data, and guide the satellite-ground integrated Internet of Things fusion system on how to allocate limited energy resources to support data regeneration and other key tasks.

[0077] By performing feature engineering on the second data, feature vectors related to the fault type and energy prediction value are extracted. Based on the extracted feature vector, the fault type label is converted into a numerical vector using unique hot encoding. The converted fault type label vector is concatenated with the energy prediction value required for data regeneration to obtain a fused feature vector. According to the dimension of the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined. The number of hidden layers and the number of nodes in each layer are set according to the rule of thumb. The number of nodes in the output layer is set to be the same as the dimension of the energy scheduling decision variable. The weight matrix and bias vector of the multilayer perceptron neural network are initialized using a random initialization method. According to the fused feature vector, the output value of the multilayer perceptron neural network is calculated by forward propagation as the prediction value of the decision variable of the energy scheduling model. According to the optimization objective function of the energy scheduling model, the loss function value under the current prediction value is calculated. The gradient of the loss function to the weight matrix and bias vector of the multilayer perceptron neural network is calculated using a back propagation algorithm. According to the calculated gradient, the weight matrix and bias vector of the multilayer perceptron neural network are updated using a stochastic gradient descent method. Repeat the process of forward propagation, loss function calculation, back propagation and parameter update until the loss function value converges or the preset number of iterations is reached. The trained multi-layer perceptron neural network is used as an energy scheduling model to predict the optimal energy scheduling decision variables based on the fault type label and the predicted value of energy required for data regeneration.

[0078] In this embodiment, the energy scheduling model constructed by the multi-layer perceptron neural network can intelligently allocate energy resources according to the fault type of the data and the regeneration energy demand, thereby improving the energy utilization efficiency and ensuring the timely recovery and processing of key data.

[0079] In some embodiments, the energy scheduling model is constructed using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of the energy required for data regeneration, specifically including:

[0080] Extracting a first feature vector by performing feature engineering processing on the second data;

[0081] According to the first feature vector, the fault type label is converted into a numerical vector by using the unique hot encoding of the second data to obtain a fault type label vector;

[0082] Concatenate the fault type label vector and the predicted value of energy required for data regeneration to obtain a fused feature vector;

[0083] Use random initialization method to initialize the weight matrix and bias vector of the multilayer perceptron neural network;

[0084] According to the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined, and the output value of the multilayer perceptron neural network is calculated by forward propagation;

[0085] The back propagation algorithm is used to calculate the gradient of the loss function with respect to the weight matrix and bias vector of the multi-layer perceptron neural network.

[0086] According to the gradient, the stochastic gradient descent method is used to update the weight matrix and bias vector of the multilayer perceptron neural network, and the process of forward propagation, loss function calculation, back propagation and parameter update is repeated until the value of the loss function converges or reaches a preset number of iterations to obtain an energy scheduling model.

[0087] In this embodiment, based on the second data, a first feature vector capable of characterizing the characteristics of the data is extracted through a feature engineering processing method to obtain a feature representation with business significance; the fault type label is converted from text form to a numerical vector representation using a unique hot encoding method for the second data to obtain a fault type label vector that can be used for model training; the fault type label vector and the predicted value of energy required for data regeneration are concatenated to obtain a fused feature vector that combines information on both the fault type and the energy consumption prediction through feature combination.

[0088] According to the dimension of the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined, and the weight matrix and bias vector of the multilayer perceptron neural network are initialized by a random initialization method to obtain the initialized neural network model parameters. The fused feature vector is input into the input layer of the multilayer perceptron neural network, and the weighted sum of neurons and the output of the activation function are calculated layer by layer through forward propagation calculation, and finally the output value of the multilayer perceptron neural network is obtained as the prediction result of the energy scheduling model. According to the prediction result of the energy scheduling model and the actual energy scheduling data, the prediction error is calculated by the loss function, and the gradient of the loss function to the weight matrix and bias vector of the multilayer perceptron neural network is calculated layer by layer through the back propagation algorithm. According to the calculated gradient, the weight matrix and bias vector of the multilayer perceptron neural network are updated by the random gradient descent method to obtain the updated neural network model parameters. Repeat the execution, and continuously perform the iterative process of forward propagation, loss function calculation, back propagation and parameter update until the value of the loss function converges or reaches the preset number of iterations, and judge whether the model training is completed. When the model training is completed, the final multilayer perceptron neural network model is obtained as the energy scheduling model, which is used to predict and optimize the energy scheduling of new input data.

[0089] For example, firstly, feature engineering is performed on the second data to extract feature vectors related to the fault type and energy prediction value, such as the time of occurrence of the fault, the duration of the fault, the type of faulty equipment, the severity of the fault, etc. Then, the fault type label is converted into a numerical vector using unique hot encoding, for example, "equipment failure" is encoded as [1, 0, 0], "network failure" is encoded as [0, 1, 0], and "software failure" is encoded as [0, 0, 1]. Then, the converted fault type label vector is concatenated with the energy prediction value required for data regeneration to obtain a fused feature vector. Assuming that the dimension of the fault type label vector is 3 and the energy prediction value is a real number, the dimension of the fused feature vector is 4. According to the dimension of the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined to be 4. Using the rule of thumb, the number of hidden layers is set to 2, the number of nodes in the first layer is 8, and the number of nodes in the second layer is 4. The number of nodes in the output layer is set to be the same as the dimension of the energy scheduling decision variable. Assuming that the energy scheduling decision variables include the output of each generator set, the charging and discharging power of each energy storage device, the power demand of the load, etc., a total of 10 variables, the number of nodes in the output layer is 10. The weight matrix and bias vector of the multilayer perceptron neural network are initialized using the random initialization method of Gaussian distribution. According to the fused feature vector, the output value of the multilayer perceptron neural network is calculated by forward propagation as the predicted value of the decision variable of the energy scheduling model. According to the optimization objective function of the energy scheduling model, such as minimizing the power generation cost and maximizing the environmental benefit, the loss function value under the current predicted value is calculated. The back propagation algorithm is used to calculate the gradient of the loss function to the weight matrix and bias vector of the multilayer perceptron neural network based on the chain rule. According to the calculated gradient, the Adam optimization algorithm is used to update the weight matrix and bias vector of the multilayer perceptron neural network, and the learning rate is set to 01, the momentum factor is set to 9, and the number of iterations is set to 1000. The above training process is repeated until the loss function value is less than 0.01 or 1000 iterations are reached. Finally, the trained multi-layer perceptron neural network is used as the energy scheduling model. By inputting the fault type label and the predicted value of the energy required for data regeneration, the optimal energy scheduling decision variables can be predicted to achieve intelligent and efficient energy management.

[0090] S104, regenerating the second data according to the data regeneration processing scheme to obtain a data regeneration result, and based on the data regeneration result, using a support vector machine fault diagnosis algorithm to identify the fault type and fault location of the satellite-ground integrated Internet of Things terminal.

[0091] In this embodiment, the data regeneration processing scheme adopts an autoencoder neural network algorithm, and removes noise and redundant information in the data by performing feature extraction and dimension reduction on the second data, and reconstructs a purer and more complete data regeneration result. Based on the data regeneration result, a support vector machine fault diagnosis algorithm is used to identify the fault type and fault location of the satellite-ground integrated Internet of Things terminal. The data regeneration result is used as the input of the support vector machine algorithm, and the data regeneration result is mapped to a feature vector and classified and decided to obtain the fault type discrimination result and the fault location location result. If the fault type discrimination result shows that there is a fault, the specific fault type is determined according to the fault type discrimination result; at the same time, if the fault location location result shows that the fault location is clear, the specific fault location is determined according to the fault location location result. By integrating the fault type discrimination result and the fault location location result, the comprehensive fault diagnosis result of the satellite-ground integrated Internet of Things terminal is obtained, which fully reflects the fault condition of the satellite-ground integrated Internet of Things terminal and provides an accurate diagnosis basis for subsequent fault repair and recovery. The comprehensive fault diagnosis result includes fault type information and fault location information. By analyzing the comprehensive fault diagnosis result, the fault condition of the satellite-ground integrated Internet of Things terminal can be fully grasped, the severity of the fault can be judged, and a targeted fault handling plan can be formulated. According to the comprehensive fault diagnosis results, combined with the system architecture and functional modules of the satellite-ground integrated IoT terminal, a fine-grained analysis of the fault type and fault location is performed to further clarify the root cause of the fault, find out the key influencing factors of the fault, and provide support for in-depth diagnosis of the fault. Through in-depth analysis of the fault diagnosis results, a fault recovery strategy for the satellite-ground integrated IoT terminal is formulated, and corresponding fault recovery measures are taken for different fault types and fault locations, such as restarting the fault module, reconfiguring system parameters, updating software versions, etc., to achieve rapid recovery of the satellite-ground integrated IoT terminal.

[0092] S105, calculating the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determining the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location.

[0093] In this embodiment, the energy scheduling model is used to calculate the energy consumption value required for terminal fault self-healing. Among them, the energy scheduling model is an energy optimization model based on the Markov decision process, and the optimal energy scheduling strategy is solved by the dynamic programming algorithm. According to the calculated energy consumption value required for fault self-healing, it is judged whether the current remaining energy of the terminal meets the energy required for self-healing. If it meets, the decision tree algorithm is used to generate a specific fault self-healing plan according to the fault type and fault location information. The decision tree algorithm establishes a mapping relationship between fault characteristics and self-healing operations by learning historical fault self-healing cases, so as to obtain the optimal self-healing operation combination for different fault types and locations. If the current remaining energy of the terminal does not meet the energy required for self-healing, an energy replenishment request is triggered to request other terminals or energy replenishment stations in the satellite-ground integrated network to provide energy support. After obtaining enough energy, the previous step is re-executed to generate a specific fault self-healing plan. After the fault self-healing plan is generated, the self-healing plan is transmitted to the corresponding fault terminal, and the process of the terminal performing the self-healing operation is monitored. Through the abnormal detection algorithm deployed on the terminal, various performance parameters of the terminal are analyzed in real time to determine whether the fault is repaired successfully. The anomaly detection algorithm uses a single classifier model based on a support vector machine. By training the terminal performance data in normal state, a judgment boundary of normal working state is constructed. If the terminal performance parameters exceed this boundary, it is judged that the fault has not been repaired. If the fault is repaired successfully, the terminal status information is updated to complete the self-healing process; if the fault is not repaired, it returns to the first step, re-acquires the terminal status and performs diagnosis and analysis, and generates a new self-healing plan until the fault is repaired successfully. During the entire self-healing process, the fault diagnosis, energy scheduling and self-healing decision models are updated and optimized in real time using incremental learning, so that they can adapt to the ever-changing fault scenarios and improve the accuracy and efficiency of self-healing.

[0094] For example, a terminal has a communication module failure, the fault location is in the third quadrant of the PCB board, and the current remaining battery power of the terminal is 500mAh. The system inputs this information into the energy scheduling model based on the Markov decision process. The model uses the dynamic programming algorithm to calculate that the energy required to repair the fault is 800mAh. Since the terminal is currently low on power, the system sends an energy replenishment request to the nearest refueling station, requesting 300mAh of energy. After receiving the request, the refueling station transmits energy to the faulty terminal through wireless charging technology. After the energy transmission is completed, the system uses the decision tree algorithm to find the corresponding self-healing solution in historical cases based on the characteristics of the communication module failure and the third quadrant position of the PCB board, and generates the optimal repair operation sequence for this fault, which mainly includes steps such as restarting the communication module and updating the firmware version, and transmits the self-healing solution to the faulty terminal. After receiving the self-healing solution, the faulty terminal starts to perform a series of self-healing operations. At the same time, the system collects data from various sensors of the terminal in real time, extracts 15 performance parameters such as communication delay and packet loss rate, and inputs them into the trained support vector machine single classifier for fault detection. If the classifier determines that the terminal is in normal condition, the self-healing process is completed and the terminal status information is updated; if the classifier determines that the fault has not been repaired, the system will re-diagnose and analyze the terminal status and adjust the self-healing plan based on the analysis results, such as increasing the antenna transmission power, switching to the backup communication link, etc., until the terminal resumes normal operation. Finally, the system uses the data collected during this self-healing process to optimize and update the fault diagnosis model, energy scheduling strategy, and self-healing decision tree through incremental learning, so that the system can respond to various future fault conditions more intelligently and efficiently.

[0095] In some embodiments, determining the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location specifically includes:

[0096] Acquire historical fault feature data and historical fault self-healing data, establish a mapping relationship between the historical fault feature data and the historical fault self-healing data, and use a decision tree algorithm to establish a fault self-healing solution generation model, wherein the fault self-healing solution generation model is used to obtain an optimal self-healing operation combination for different fault types and fault locations;

[0097] Determine whether the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing;

[0098] If it is satisfied, the current optimal fault self-healing solution is directly generated for the satellite-ground integrated Internet of Things terminal according to the fault self-healing solution generation model; otherwise, an energy replenishment request is triggered until the current remaining energy of the satellite-ground integrated Internet of Things terminal meets the energy consumption value required for fault self-healing.

[0099] In this embodiment, historical fault feature data and historical fault self-healing data are obtained, including attributes such as fault occurrence time, fault type, fault location, and self-healing operation. Through data preprocessing, the acquired historical fault feature data and historical fault self-healing data are cleaned, normalized, and other processes are performed to obtain a standardized data set. Data mining technology is used to analyze the association rules and patterns between historical fault feature data and historical fault self-healing data, and a mapping relationship between fault features and self-healing operations is established. According to the established mapping relationship, a decision tree algorithm is used to train a fault self-healing solution generation model. The decision tree uses fault features as nodes, selects the optimal splitting attributes through indicators such as information gain, and recursively constructs a binary tree model. The trained decision tree model is evaluated by methods such as cross-validation, optimizes model parameters, and improves the accuracy and generalization ability of the fault self-healing solution. The optimized decision tree model is saved as a fault self-healing solution generation model, which is used to quickly generate the optimal self-healing operation combination for new faults. When a system fault occurs, characteristic information of the current fault is obtained, including attributes such as fault type and fault location. The fault characteristics are input into the trained fault self-healing solution generation model, and the model automatically infers the optimal self-healing operation combination for the fault based on its internal decision rules.

[0100] The current remaining energy value of the satellite-ground integrated IoT terminal is obtained, and the remaining energy of the satellite-ground integrated IoT terminal is monitored in real time through the energy detection module to obtain the current remaining energy value. According to the fault self-healing scheme model, the energy consumption estimation algorithm is used to calculate the energy consumption value required for executing the current fault self-healing scheme, and the energy consumption value required for fault self-healing is obtained. The current remaining energy value of the satellite-ground integrated IoT terminal is compared with the energy consumption value required for fault self-healing to determine whether the current remaining energy meets the energy demand for fault self-healing. If the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing, the optimal fault self-healing scheme under the current situation is generated according to the fault self-healing scheme generation model, the optimal decision algorithm is used, and the real-time state parameters of the satellite-ground integrated IoT terminal are combined, and the scheme is executed to realize the fault self-healing of the satellite-ground integrated IoT terminal. If the current remaining energy of the satellite-ground integrated IoT terminal does not meet the energy consumption value required for fault self-healing, an energy replenishment request is triggered, and energy is replenished for the satellite-ground integrated IoT terminal through the energy replenishment system. After the energy replenishment is completed, the current remaining energy value of the satellite-ground integrated IoT terminal is obtained again to determine whether it meets the energy consumption value required for fault self-healing, until the current remaining energy meets the conditions. When the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing, it returns to continue to generate the model according to the fault self-healing solution, generates and executes the optimal fault self-healing solution, and ensures that the satellite-ground integrated IoT terminal completes the fault self-healing process with sufficient energy, ensuring the reliable operation of the system.

[0101] S106, obtaining the energy supply status of each satellite-ground integrated Internet of Things terminal, and dynamically adjusting the energy allocation scheme between each satellite-ground integrated Internet of Things terminal by using a dynamic programming algorithm according to the fault self-healing scheme and the energy supply status.

[0102] In this embodiment, the remaining energy and energy consumption rate of each satellite-ground integrated IoT terminal are obtained through the energy detection module to obtain the energy supply of each terminal. According to the energy supply of each terminal obtained, an energy distribution optimization model is constructed to convert the energy distribution problem into a dynamic programming problem. In the energy distribution optimization model, the state variable is defined as the remaining energy of each terminal, and the decision variable is the energy distribution plan between each terminal. The energy distribution optimization model is solved by a dynamic programming algorithm to obtain the optimal energy distribution plan between each satellite-ground integrated IoT terminal. The core of the dynamic programming algorithm is to use the optimal substructure property to decompose the problem into several sub-problems, and then obtain the optimal solution of the original problem by solving the optimal solution of the sub-problem. According to the optimal energy distribution plan obtained by the dynamic programming algorithm, the energy supply module is controlled to supply energy to each satellite-ground integrated IoT terminal. The energy supply module distributes energy to each terminal according to the optimal energy distribution plan to ensure that the energy supply of each terminal meets its working needs. During the energy distribution and supply process, the energy consumption of each satellite-ground integrated IoT terminal is continuously monitored. If the energy consumption of a terminal is found to be abnormal, the energy distribution plan is adjusted in time to ensure the stability of the energy supply of each terminal. Through the above steps, the dynamic programming algorithm is used to dynamically optimize the energy distribution between the satellite-ground integrated IoT terminals to ensure that the energy supply of each terminal meets its working needs and realize the efficient operation of the IoT system. In the process of energy optimization and allocation, each step is closely related, and the output of the previous step is used as the input of the next step to form a complete energy optimization closed loop.

[0103] For example, the remaining energy of 10 satellite-ground integrated IoT terminals obtained by the energy detection module is 1000mAh, 800mAh, 1200mAh, 900mAh, 1100mAh, 950mAh, 1050mAh, 850mAh, 1150mAh and 1000mAh, and the energy consumption rate is 100mAh / h, 120mAh / h, 90mAh / h, 110mAh / h, 95mAh / h, 105mAh / h, 100mAh / h, 115mAh / h, 85mAh / h and 100mAh / h. According to the energy supply of each terminal, an energy allocation optimization model is constructed, and the state variable is defined as the remaining energy of each terminal, and the decision variable is defined as the energy allocation plan between the terminals. The dynamic programming algorithm is used to solve the energy allocation optimization model, and the state transfer equation is set to f(i, j) = max{f(i-1, j), f(i-1, j-wi) + vi}, where f(i, j) represents the maximum energy utilization efficiency of the first i terminals when the total energy is j, wi represents the energy consumption rate of the i-th terminal, and vi represents the energy utilization efficiency of the i-th terminal. Through bottom-up recursive calculation, the optimal energy allocation scheme between each satellite-ground integrated IoT terminal is obtained, which is 200mAh, 150mAh, 250mAh, 100mAh, 200mAh, 150mAh, 100mAh, 200mAh, 250mAh and 150mAh respectively. According to the optimal energy allocation scheme, the energy supply module is controlled to allocate energy to each terminal to ensure that the energy supply of each terminal meets its working needs. During the energy distribution and supply process, the energy consumption of each satellite-ground integrated IoT terminal is continuously monitored. If the energy consumption of a terminal is found to be abnormal, such as the energy consumption rate of the fifth terminal suddenly increases to 150mAh / h, the energy distribution plan is adjusted in time to increase the energy distribution of the fifth terminal to 250mAh to ensure the stability of the energy supply of each terminal. The energy distribution between the satellite-ground integrated IoT terminals is dynamically optimized through the dynamic programming algorithm to achieve efficient operation of the IoT system.

[0104] Reference Figure 2 An embodiment of the present invention provides a satellite-ground integrated Internet of Things communication processing system 2 based on multi-network integration, and the system 2 specifically includes:

[0105] The first processing module 201 is used to obtain the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite, switch to a backup network mode, where the backup network mode includes a cellular network, a LoRa network, and a 6G network, and at the same time, perform fault type label identification on the first data to obtain second data;

[0106] The second processing module 202 is used to obtain environmental data between the ground control center and the low-orbit satellite, determine a predicted value of energy required for data regeneration after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determine a data regeneration processing scheme for data regeneration of the second data in the ground control center according to the predicted value of energy required for data regeneration;

[0107] The third processing module 203 is used to construct an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration;

[0108] The fourth processing module 204 is used to regenerate the second data according to the data regeneration processing scheme to obtain a data regeneration result, and based on the data regeneration result, use a support vector machine fault diagnosis algorithm to identify the fault type and fault location of the satellite-ground integrated Internet of Things terminal;

[0109] The fifth processing module 205 is used to calculate the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determine the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location;

[0110] The sixth processing module 206 is used to obtain the energy supply status of each satellite-ground integrated Internet of Things terminal, and dynamically adjust the energy allocation plan between each satellite-ground integrated Internet of Things terminal by using a dynamic programming algorithm according to the fault self-healing plan and the energy supply status.

[0111] It is understandable that if Figure 1 The contents of the embodiment of the satellite-ground integrated Internet of Things communication processing method based on multi-network integration shown in the figure are all applicable to the embodiment of the satellite-ground integrated Internet of Things communication processing system based on multi-network integration. The functions specifically implemented by the embodiment of the satellite-ground integrated Internet of Things communication processing system based on multi-network integration are similar to those in the embodiment of the satellite-ground integrated Internet of Things communication processing system. Figure 1 The embodiment of the satellite-ground integrated Internet of Things communication processing method based on multi-network integration shown in FIG. Figure 1 The beneficial effects achieved by the embodiment of the satellite-ground integrated Internet of Things communication processing method based on multi-network integration shown are also the same.

[0112] It should be noted that the information interaction, execution process and other contents between the above-mentioned systems are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0113] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0114] Reference Figure 3 The embodiment of the present invention further provides a computer device 3, comprising: a memory 302 and a processor 301 and a computer program 303 stored in the memory 302. When the computer program 303 is executed on the processor 301, a satellite-ground integrated Internet of Things communication processing method based on multi-network integration as described in any one of the above methods is implemented.

[0115] The computer device 3 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will appreciate that Figure 3 It is only an example of computer device 3 and does not constitute a limitation on computer device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0116] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0117] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device 3. Further, the memory 302 may also include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0118] An embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for processing satellite-ground integrated Internet of Things communications based on multi-network integration as described in any one of the above methods is implemented.

[0119] In this embodiment, if the integrated unit is implemented in the form of 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 present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0120] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0121] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example 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 performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0122] In the embodiments disclosed in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0123] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Claims

1. A satellite-ground integrated Internet of Things communication processing method based on multi-network integration, characterized in that: The method specifically comprises: Acquire the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting first data to the low-orbit satellite, switch to a backup network mode, where the backup network mode includes a cellular network, a LoRa network, and a 6G network, and perform fault type label identification on the first data to obtain second data; The step of performing fault type label identification on the first data to obtain second data specifically includes: Obtain historical data transmitted by the satellite-ground integrated IoT terminal to the low-orbit satellite, add a fault type label to each abnormal data in the historical data, and form a fault diagnosis data set; Extracting key features from the fault diagnosis data set to obtain feature vectors, training the feature vectors using a support vector machine algorithm to obtain a fault diagnosis feature space representation; After associating the fault diagnosis feature space representation with each fault type, an associated data set is formed, and the associated data set is used as input to train a decision tree model to establish a fault type diagnosis model; Performing fault type label identification on the first data according to the fault type diagnosis model to obtain second data; Acquire environmental data between a ground control center and a low-orbit satellite, determine a predicted value of energy required for data regeneration of the second data after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determine a data regeneration processing scheme for data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration; Building an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration; The energy scheduling model is constructed by using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of the energy required for data regeneration, specifically including: Extracting a first feature vector by performing feature engineering processing on the second data; According to the first feature vector, the fault type label is converted into a numerical vector by using the unique hot encoding of the second data to obtain a fault type label vector; Concatenate the fault type label vector and the predicted value of energy required for data regeneration to obtain a fused feature vector; Use random initialization method to initialize the weight matrix and bias vector of the multilayer perceptron neural network; According to the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined, and the output value of the multilayer perceptron neural network is calculated by forward propagation; The back propagation algorithm is used to calculate the gradient of the loss function with respect to the weight matrix and bias vector of the multi-layer perceptron neural network. According to the gradient, the weight matrix and bias vector of the multilayer perceptron neural network are updated by using the stochastic gradient descent method, and the process of forward propagation, loss function calculation, back propagation and parameter update is repeatedly performed until the value of the loss function converges or reaches a preset number of iterations, thereby obtaining an energy scheduling model; Regenerate the second data according to the data regeneration processing scheme to obtain a data regeneration result, and based on the data regeneration result, use a support vector machine fault diagnosis algorithm to identify the fault type and fault location of the satellite-ground integrated Internet of Things terminal; Calculate the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determine the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location; The method of determining the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for the fault self-healing, the fault type and the fault location specifically includes: Acquire historical fault feature data and historical fault self-healing data, establish a mapping relationship between the historical fault feature data and the historical fault self-healing data, and use a decision tree algorithm to establish a fault self-healing solution generation model, wherein the fault self-healing solution generation model is used to obtain an optimal self-healing operation combination for different fault types and fault locations; Determine whether the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing; If satisfied, the optimal fault self-healing solution is directly generated for the satellite-ground integrated IoT terminal according to the fault self-healing solution generation model; otherwise, an energy replenishment request is triggered until the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing; The energy supply conditions of each satellite-ground integrated Internet of Things terminal are obtained, and according to the fault self-healing scheme and the energy supply conditions, a dynamic programming algorithm is used to dynamically adjust the energy allocation scheme between each satellite-ground integrated Internet of Things terminal.

2. The method according to claim 1, characterized in that When an abnormality occurs during the process of transmitting the first data from the satellite-ground integrated Internet of Things terminal to the low-orbit satellite, switching to the backup network mode specifically includes: Determine the coordinate information of the satellite-ground integrated Internet of Things terminal according to the real-time position, insert the coordinate information into the data packet header of the first data, and transmit it to the ground control center; During the transmission of the first data, network status parameters are collected in real time, and according to the network status parameters, a hidden Markov model is used to model the network status between the satellite-ground integrated Internet of Things terminal and the low-orbit satellite to obtain a network status hidden Markov model; The Viterbi algorithm is used to solve the network state hidden Markov model to obtain the optimal state transfer path of the network state. According to the optimal state transfer path of the network state, it is determined whether an abnormality occurs in the process of the current satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite. If it is determined to be abnormal, the backup network mode is used to continue the transmission of the first data.

3. The method according to claim 1, characterized in that: The step of determining the predicted value of energy required for data regeneration after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data specifically includes: Acquire historical transmission loss degree data of second data transmitted from the low-orbit satellite to the ground control center, and determine a data set of energy value required for data regeneration according to the historical transmission loss degree data; Associating the environmental data with the data set of energy value required for data regeneration, and inputting them into a support vector machine regression model or a long short-term memory neural network model for training to obtain a prediction model of energy required for data regeneration; The data regeneration required energy prediction value after the current second data is transmitted from the low-orbit satellite to the ground control center is determined according to the data regeneration required energy prediction model.

4. The method according to claim 1, characterized in that: The step of determining the data regeneration processing scheme for performing data regeneration on the second data at the ground control center according to the predicted value of energy required for data regeneration specifically includes: Taking the predicted value of energy required for data regeneration as a constraint condition, and taking the quality requirement after data regeneration and the efficiency requirement during efficiency regeneration as an objective function; According to the constraint conditions and the objective function, a multi-objective optimization algorithm is used to calculate and obtain a data regeneration processing plan for regenerating the second data at the ground control center.

5. A satellite-ground integrated Internet of Things communication processing system based on multi-network integration, characterized in that: The system specifically comprises: A first processing module is used to obtain the real-time position of the satellite-ground integrated Internet of Things terminal, and when an abnormality occurs during the process of the satellite-ground integrated Internet of Things terminal transmitting the first data to the low-orbit satellite, switch to a backup network mode, where the backup network mode includes a cellular network, a LoRa network, and a 6G network, and at the same time, perform fault type label identification on the first data to obtain second data; The step of performing fault type label identification on the first data to obtain second data specifically includes: Obtain historical data transmitted by the satellite-ground integrated IoT terminal to the low-orbit satellite, add a fault type label to each abnormal data in the historical data, and form a fault diagnosis data set; Extracting key features from the fault diagnosis data set to obtain feature vectors, training the feature vectors using a support vector machine algorithm to obtain a fault diagnosis feature space representation; After associating the fault diagnosis feature space representation with each fault type, an associated data set is formed, and the associated data set is used as input to train a decision tree model to establish a fault type diagnosis model; Performing fault type label identification on the first data according to the fault type diagnosis model to obtain second data; a second processing module, configured to obtain environmental data between the ground control center and the low-orbit satellite, determine a predicted value of energy required for data regeneration of the second data after the second data is transmitted from the low-orbit satellite to the ground control center through the environmental data, and determine a data regeneration processing scheme for data regeneration of the second data at the ground control center according to the predicted value of energy required for data regeneration; A third processing module is used to construct an energy scheduling model using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of energy required for data regeneration; The energy scheduling model is constructed by using a multi-layer perceptron neural network according to the fault type label of the second data and the predicted value of the energy required for data regeneration, specifically including: Extracting a first feature vector by performing feature engineering processing on the second data; According to the first feature vector, the fault type label is converted into a numerical vector by using the unique hot encoding of the second data to obtain a fault type label vector; Concatenate the fault type label vector and the predicted value of energy required for data regeneration to obtain a fused feature vector; Use random initialization method to initialize the weight matrix and bias vector of the multilayer perceptron neural network; According to the fused feature vector, the number of input layer nodes of the multilayer perceptron neural network is determined, and the output value of the multilayer perceptron neural network is calculated by forward propagation; The back propagation algorithm is used to calculate the gradient of the loss function with respect to the weight matrix and bias vector of the multi-layer perceptron neural network. According to the gradient, the weight matrix and bias vector of the multilayer perceptron neural network are updated by using the stochastic gradient descent method, and the process of forward propagation, loss function calculation, back propagation and parameter update is repeatedly performed until the value of the loss function converges or reaches a preset number of iterations, thereby obtaining an energy scheduling model; a fourth processing module, configured to perform data regeneration on the second data according to the data regeneration processing scheme to obtain a data regeneration result, and identify a fault type and a fault location of the satellite-ground integrated Internet of Things terminal by using a support vector machine fault diagnosis algorithm based on the data regeneration result; A fifth processing module is used to calculate the energy consumption value required for fault self-healing of the satellite-ground integrated Internet of Things terminal according to the energy scheduling model, and determine the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for fault self-healing, the fault type and the fault location; The method of determining the fault self-healing solution of the satellite-ground integrated Internet of Things terminal according to the energy consumption value required for the fault self-healing, the fault type and the fault location specifically includes: Acquire historical fault feature data and historical fault self-healing data, establish a mapping relationship between the historical fault feature data and the historical fault self-healing data, and use a decision tree algorithm to establish a fault self-healing solution generation model, wherein the fault self-healing solution generation model is used to obtain an optimal self-healing operation combination for different fault types and fault locations; Determine whether the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing; If satisfied, the optimal fault self-healing solution is directly generated for the satellite-ground integrated IoT terminal according to the fault self-healing solution generation model; otherwise, an energy replenishment request is triggered until the current remaining energy of the satellite-ground integrated IoT terminal meets the energy consumption value required for fault self-healing; The sixth processing module is used to obtain the energy supply status of each satellite-ground integrated Internet of Things terminal, and dynamically adjust the energy allocation plan between each satellite-ground integrated Internet of Things terminal according to the fault self-healing plan and the energy supply status by using a dynamic programming algorithm.

Images