A remote real-time control system for industrial internet

Abstract

The application discloses a kind of remote real-time control systems for industrial internet, specifically relates to remote control technical field, including: adaptive network delay compensation module, multi-path redundant transmission module and intelligent data compression and optimization module.Adaptive network delay compensation module monitors network delay by high-frequency heartbeat packet, uses adaptive filtering algorithm to forecast, and dynamically corrects prediction model in combination with the closed-loop feedback mechanism of PI controller, to realize accurate delay compensation;The multi-path redundant transmission module dynamically allocates data packets based on real-time calculation of sending cost by establishing multiple transmission paths, to realize intelligent load balancing and automatic switching of fault path.The application effectively solves the problem of poor real-time performance and low reliability caused by network delay, jitter and bandwidth limitation in industrial internet remote control, significantly improves the comprehensive performance of the system.

Description

Technical Field

[0001] This invention relates to the field of remote control technology, and more specifically, to a remote real-time control system for the Industrial Internet. Background Technology

[0002] In the field of the Industrial Internet, remote real-time control systems are the core of realizing intelligent manufacturing and unmanned operation and maintenance. These systems transmit commands from the control terminal to remote industrial equipment via public networks and collect equipment status data in real time. However, the unreliability and heterogeneity of networks pose significant challenges to remote control. Traditional control systems are mostly designed based on local area networks (LANs), and their stable, low-latency, high-bandwidth environment is difficult to replicate in complex wide area networks (WANs). Inherent latency, jitter, and packet loss in network transmission can directly lead to control commands not arriving on time or equipment status information not being updated promptly, severely affecting the system's control accuracy and stability, and potentially even causing production safety accidents.

[0003] To address these challenges, existing technologies have proposed several solutions. For example, using a fixed delay compensation value to send instructions in advance, or configuring a single redundant network path for critical data. At the data level, common lossless or lossy compression algorithms are generally employed to reduce network load. However, these methods all have significant limitations: static compensation mechanisms cannot adapt to dynamically changing network conditions, often resulting in overcompensation or undercompensation; simple path redundancy lacks intelligent scheduling and cannot fully utilize the bandwidth aggregation and fault tolerance advantages of multi-path systems; and common compression algorithms struggle to achieve a good balance between compression ratio, processing speed, and the preservation of the unique structure of industrial control data.

[0004] Therefore, existing technologies lack a comprehensive solution that can coordinate network latency, transmission reliability, and data bandwidth limitations. Most systems only address single issues, resulting in poor overall performance when facing the complex and ever-changing remote control scenarios in the Industrial Internet, where real-time performance and reliability cannot be guaranteed simultaneously. Developing an integrated control system that can adapt to network fluctuations and possess intelligent transmission scheduling and efficient data optimization capabilities has become a pressing technical challenge in this field. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a remote real-time control system for the Industrial Internet.

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

[0007] A remote real-time control system for the Industrial Internet includes the following modules:

[0008] The adaptive network delay compensation module is used to monitor network delay through high-frequency heartbeat packets and predict delay trends using an adaptive filtering algorithm; control commands are sent in advance based on the prediction results, and a closed-loop feedback mechanism is introduced to dynamically correct the prediction model.

[0009] The multi-path redundancy transmission module is used to establish and evaluate the performance of multiple transmission paths. By dynamically calculating the transmission cost, it allocates data packets to the optimal path and automatically switches paths and retransmits when transmission fails, thereby achieving load balancing and reliable transmission.

[0010] The intelligent data compression and optimization module is used to select the appropriate compression encoding method according to the characteristics of data types. It dynamically searches for the best compression parameters through a multi-objective optimization function to achieve a balance between compression ratio, processing time and data quality, and adopts differentiated compression strategies for data with different priorities.

[0011] Specifically, the execution process of the adaptive network latency compensation module is as follows:

[0012] First, time synchronization units are deployed at both the control end and the device end to calculate the current network latency value through high-frequency heartbeat packet exchange and record historical latency data to form a latency sequence.

[0013] Secondly, an adaptive filtering algorithm is used to smooth the delayed sequence, remove noise and jitter, and predict the delay trend of the next time period.

[0014] Then, the sending time of the control command is adjusted according to the predicted delay, and the command is sent in advance to offset the effect of the delay. At the same time, a dynamic buffer is set to store the command and release it according to the actual arrival time.

[0015] A closed-loop feedback control mechanism is introduced, which uses the error between the predicted value and the actual measured value as a continuous feedback signal to correct the prediction model itself.

[0016] Finally, monitor latency changes in real time, and if the latency exceeds the threshold, trigger an alarm and switch to a backup control strategy.

[0017] Specifically, the closed-loop feedback control mechanism includes:

[0018] Error calculation: In each control cycle, when the latest actual network latency measurement is obtained via heartbeat packets... Then, calculate the corresponding predicted value. The error between them ,in, This is the prediction error at the current moment;

[0019] Feedback control quantity calculation: A digital PI (proportional-integral) controller is used to process this error signal; the PI controller generates the model correction quantity based on the current error (proportional term) and the cumulative error history (integral term);

[0020] Scale term P: Provides the current error A proportional response, designed to reduce errors immediately;

[0021] Integral term I: Cumulative past errors , used to eliminate steady-state errors;

[0022] Control quantity The calculation formula is:

[0023]

[0024] in, For production volume, This is the cumulative sum of errors from time 0 to t. It is proportional gain. Integral gain;

[0025] Online calibration of model parameters: The calculated control inputs are then calibrated online. It applies to the existing adaptive filtering prediction model;

[0026] When the weight vector of the original adaptive filter is The update rule can then be revised as follows:

[0027]

[0028] in, Let be the value of the weight vector of the adaptive filter at time t. It is the step size of the original filter. It is the input vector. It is a mixing factor less than 1;

[0029] Alternatively, directly correct the predicted output: ,in These are the new predicted values ​​after feedback calibration. These are the original predicted values;

[0030] Monitoring and Reset: Continuously monitor errors The absolute value or moving average; when the error continues to exceed the preset threshold for a preset time, it is determined that the network has undergone drastic and fundamental changes, and the prediction model is deemed inapplicable at this time;

[0031] The automatic reset mechanism clears the error accumulation term of the PI controller and restores the filter weights to their default values, allowing the model to start learning and adapting anew on a new basis.

[0032] Specifically, the execution process of the multi-path redundancy transmission module is as follows:

[0033] Establish multiple network connections at the control and device ends, including wired, wireless, and mobile network paths, and evaluate the bandwidth, latency, and packet loss rate of each path;

[0034] The control commands are divided into multiple data packets, and a sequence number and checksum are added to each data packet to perform dynamic data packet allocation and load balancing.

[0035] The device receives data packets, reassembles instructions according to the sequence number, and verifies data integrity using a checksum. If a transmission path fails or the delay exceeds a threshold, it automatically switches to another path and retransmits the lost data packets.

[0036] Regularly optimize path selection strategies and dynamically adjust transmission paths based on real-time network performance data to ensure the lowest latency and highest reliability.

[0037] Specifically, the process of performing dynamic packet allocation and load balancing is as follows:

[0038] Continuously monitor key performance indicators for each transmission path, including:

[0039] Available bandwidth Current load Current delay , Indicates the first Path;

[0040] For each data packet to be sent, calculate the sending cost on each path. The formula is as follows: ,in, It is a dynamic load factor used to balance the weights of load and latency.

[0041] Select sending cost The shortest path is used to send the current data packet; when the cost difference between multiple paths is less than a preset threshold, a round-robin strategy is used to distribute data packets among these high-quality paths to avoid overloading a single path.

[0042] Dynamically adjust load factor based on network congestion level. :

[0043] Automatically increase [the number of devices] when a network average latency increase of more than 20% or a packet loss rate increase is detected. A value of λ+0.1 makes the system focus more on load balancing; when network conditions meet the criteria, the value is reduced. The value is λ−0.05, which makes the system pay more attention to transmission delay;

[0044] The cost of all paths is reassessed every 100ms, and the load on a particular path is detected. When the load exceeds 50% of the average load of other paths, some of the data packets to be sent will be automatically redistributed to the less loaded paths.

[0045] Specifically, the execution process of the intelligent data compression and optimization module is as follows:

[0046] Analyze the types and characteristics of control data, and use differential coding compression for periodic data and dictionary coding compression for non-periodic data;

[0047] Machine learning algorithms are used to identify redundant patterns in the data and dynamically adjust the compression level to maximize the compression ratio while ensuring data integrity; a balance optimization mechanism between compression quality and efficiency is introduced.

[0048] Priority tags are added to compressed data before transmission. For example, emergency control instructions use low-latency compression algorithms, while historical logs use high-compression-ratio algorithms.

[0049] The receiving end decompresses the data and verifies its consistency. If decompression fails, it requests retransmission and adaptively adjusts the compression strategy based on network conditions.

[0050] Specifically, the process of the optimization mechanism for balancing compression quality and efficiency is as follows:

[0051] Real-time monitoring of three key performance indicators:

[0052] Compression ratio ;

[0053] Processing time (Time from start to finish of compression);

[0054] Data quality ;

[0055] Construct a comprehensive optimization objective function: ,in: It is a weighted index. To comprehensively optimize the objective function value;

[0056] Parameter space search and optimization:

[0057] The compression parameter space is defined as follows:

[0058] Compression level l;

[0059] Dictionary size d;

[0060] Quantization step size q;

[0061] The optimal solution is searched in the parameter space using the simulated annealing algorithm:

[0062] initial temperature The cooling coefficient c = 0.95;

[0063] Perform a neighborhood search at each temperature and accept the probability of a suboptimal solution. ,in The probability of accepting an inferior solution, The change in the objective function value is represented by T, where T is the current temperature in the simulated annealing algorithm; when the temperature drops to... Or stop when the maximum number of iterations is reached;

[0064] Dynamic weight adjustment mechanism:

[0065] Adjust the weighting index according to network conditions and application requirements:

[0066] When network bandwidth meets the requirements: increase ;

[0067] When network latency exceeds the limit: increase ;

[0068] Preset key control commands: Add ;

[0069] Weighting adjustment formula: ,in To adjust the step size,

[0070] This is the current bandwidth. This is the bandwidth threshold;

[0071] Continuously monitor the actual compression effect. If the compression performance drops by more than 10%, re-trigger the optimization search, record the historical best parameter combination, and establish a scenario-based parameter library.

[0072] The technical effects and advantages of this invention are as follows:

[0073] This invention, through an adaptive network delay compensation module, can dynamically predict and compensate for transmission delays caused by network fluctuations, significantly improving the real-time performance and accuracy of control commands, and effectively overcoming the lag problem of traditional fixed compensation strategies.

[0074] Through the intelligent load balancing and dynamic path switching mechanism of the multi-path redundant transmission module, the system achieves high reliability and stability of data transmission in complex network environments, while making full use of the aggregated bandwidth of multiple paths.

[0075] The intelligent data compression and optimization module uses a multi-objective optimization algorithm to significantly reduce transmission load while ensuring data quality. This enables the system to adaptively adjust compression strategies based on real-time network conditions, thereby comprehensively improving bandwidth utilization efficiency. Attached Figure Description

[0076] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0077] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0078] like Figure 1 As shown, a remote real-time control system module for the Industrial Internet is as follows:

[0079] The adaptive network delay compensation module monitors network delay through high-frequency heartbeat packets and predicts delay trends using an adaptive filtering algorithm. Subsequently, it sends control commands in advance based on the prediction results and introduces a closed-loop feedback mechanism to dynamically correct the prediction model, thereby offsetting the effects of delay and achieving high-precision compensation.

[0080] First, time synchronization units are deployed at both the control and device ends. These units calculate the current network latency value through high-frequency heartbeat packet exchange and record historical latency data to form a latency sequence. Second, an adaptive filtering algorithm is used to smooth the latency sequence, removing noise and jitter, and predicting the latency trend for the next time period. Then, the transmission time of control commands is adjusted based on the predicted latency, sending commands in advance to offset the impact of latency. Simultaneously, a dynamic buffer is set to store commands, which are released based on their actual arrival time.

[0081] A closed-loop feedback control mechanism is introduced, using the error between the predicted and actual measured values ​​as a continuous feedback signal to dynamically and incrementally correct the prediction model itself. This allows the model to track and adapt to the dynamic changes in network latency, thereby gradually reducing prediction bias and achieving higher accuracy compensation. The specific process is as follows:

[0082] Error Calculation: In each control cycle, when the system obtains the latest actual network latency measurement value via heartbeat packets... Then, immediately calculate its comparison with the previously predicted value. The error between them ,in, This is the prediction error at the current moment.

[0083] Feedback control quantity calculation: The system uses a digital PI (proportional-integral) controller to process this error signal. The PI controller can generate a model correction quantity based on the current error (proportional term) and the cumulative error history (integral term).

[0084] Proportional term (P): Provides the current error. A proportionally rapid response designed to immediately reduce errors. Integral term (I): Accumulated past errors. This is used to eliminate steady-state error (i.e., a long-term fixed deviation). Control quantity The calculation formula is:

[0085]

[0086] in, This is a calibration quantity (model correction quantity), generated by the PI controller. This is the cumulative sum of errors from time 0 to t (integral term). It is proportional gain. Integral gain. These two parameters need to be tuned during initialization based on the network characteristics, for example, using the Ziegler-Nichols method or empirical values;

[0087] Online calibration of model parameters: The calculated control inputs are then calibrated online. It operates on the existing adaptive filtering prediction model. Specifically, this control variable is used as a fine-tuning factor for model weights or output values.

[0088] If the weight vector of the original adaptive filter (such as an LMS filter) is The update rule can then be revised as follows:

[0089]

[0090] in, The weight vector of an adaptive filter (such as an LMS filter) is the value at time t. It is the step size of the original filter. It is the input vector. It is a mixing factor less than 1, used to control the influence strength of the feedback control quantity and prevent overcorrection.

[0091] Alternatively, the predicted output can be modified more directly: ,in These are the new predicted values ​​after feedback calibration. The original predicted value (predicted by the adaptive filtering algorithm).

[0092] Monitoring and Reset: The system continuously monitors errors. The absolute value or moving average of the error. If the error persists above a high threshold (e.g., twice the normal threshold) for a certain period, it indicates that a drastic, fundamental change may have occurred in the network (such as a main link switch), at which point the prediction model may no longer be applicable. The system will automatically trigger a reset mechanism to clear the error accumulation term of the PI controller and may restore the filter weights to their default values, allowing the model to start learning and adapting anew on a new basis.

[0093] Finally, monitor latency changes in real time. If the latency exceeds the threshold, trigger an alarm and switch to a backup control strategy, such as local autonomous mode.

[0094] The multi-path redundancy transmission module establishes and evaluates the performance of multiple transmission paths, intelligently allocates data packets to the optimal path by dynamically calculating the transmission cost, and automatically switches paths and retransmits when transmission fails, thereby achieving load balancing and reliable transmission.

[0095] First, multiple network connections are established at the control and device ends, including wired, wireless, and mobile network paths, and the bandwidth, latency, and packet loss rate of each path are evaluated. Second, control commands are segmented into multiple data packets, and a sequence number and checksum are added to each packet. Then, dynamic packet allocation and load balancing are performed, as follows:

[0096] Continuously monitor key performance indicators for each transmission path, including:

[0097] Available bandwidth Current load Current delay , Indicates the first Path;

[0098] For each data packet to be sent, calculate the sending cost on each path. The formula is as follows: ,in, It is a dynamic load factor (ranging from 0 to 1), used to balance the impact of load and latency; it also selects the transmission cost. The shortest path is used to send the current data packet. If the cost difference between multiple paths is less than a preset threshold, a round-robin strategy is used to distribute data packets among these high-quality paths to avoid overloading a single path.

[0099] Dynamically adjust load factor based on network congestion level. :

[0100] When network congestion is detected (e.g., an increase in average latency exceeding 20% ​​or an increase in packet loss rate), automatically increase... A value (e.g., λ = λ + 0.1) makes the system more focused on load balancing; when network conditions are good, reducing... Values ​​(such as λ = λ − 0.05) make the system more focused on transmission delay;

[0101] Periodically (e.g., every 100ms), reassess the cost of all paths and check for overload on any path. If the load is significantly higher than other paths (e.g., exceeding 50% of the average load), some of the packets to be sent will be automatically redistributed to the less loaded paths.

[0102] Then, the device receives data packets, reassembles instructions according to the sequence number, and verifies data integrity using a checksum. If a transmission path fails or latency is too high, it automatically switches to another path and retransmits lost data packets. Finally, the path selection strategy is periodically optimized, dynamically adjusting the transmission path based on real-time network performance data to ensure minimum latency and maximum reliability.

[0103] The intelligent data compression and optimization module selects the appropriate compression encoding method based on the characteristics of the data type. Furthermore, it dynamically searches for the optimal compression parameters through a multi-objective optimization function, achieving a balance between compression ratio, processing time, and data quality, and employs differentiated compression strategies for data of different priorities.

[0104] First, the type and characteristics of the control data are analyzed. Differential encoding compression is used for periodic data, and dictionary encoding compression is used for non-periodic data. Second, machine learning algorithms are used to identify redundancy patterns in the data and dynamically adjust the compression level to maximize the compression ratio while ensuring data integrity. A balancing optimization mechanism between compression quality and efficiency is introduced, as follows:

[0105] Real-time monitoring of three key performance indicators:

[0106] Compression ratio Processing time (Time from start to finish of compression); Data quality Construct a comprehensive optimization objective function: ,in: This is a weighting index that reflects the importance of each objective; the default setting is... ; To comprehensively optimize the objective function value;

[0107] Parameter space search and optimization: Defining the compressed parameter space includes:

[0108] Compression level l (levels 1-9); dictionary size d (32KB-1MB); quantization step size q (1-50);

[0109] The simulated annealing algorithm is used to search for the optimal solution in the parameter space: initial temperature. The cooling coefficient c = 0.95; a neighborhood search is performed at each temperature, and the probability of accepting a suboptimal solution is considered. ,in The probability of accepting an inferior solution, Let T be the change in the objective function value (the difference between the new solution and the current solution), and let T be the current temperature (in the simulated annealing algorithm); when the temperature drops to... Or stop when the maximum number of iterations is reached;

[0110] Dynamic weight adjustment mechanism:

[0111] Adjust the weighting index according to network conditions and application requirements:

[0112] When network bandwidth is sufficient: increase (Pay attention to compression ratio); When network latency is sensitive: increase (Emphasis on processing speed); Key control commands: Add (Emphasis on data quality);

[0113] Weighting adjustment formula: ,in To adjust the step size,

[0114] This is the current bandwidth. The bandwidth threshold is used; the actual compression effect is continuously monitored. If the compression performance decreases (O value decreases by more than 10%), the optimization search is retried, the historical best parameter combination is recorded, and a scenario-based parameter library is established.

[0115] Then, priority tags are added to the compressed data before transmission. High-priority data (such as emergency control instructions) uses a low-latency compression algorithm, while low-priority data (such as historical logs) uses a high compression ratio algorithm. Finally, the data is decompressed at the receiving end, and data consistency is verified. If decompression fails, a retransmission is requested, and the compression strategy is adaptively adjusted according to network conditions.

[0116] The above formulas are all dimensionless calculations. Dimensionless calculations can be performed using various methods such as standardization, which will not be elaborated here. The formulas are derived from software simulations based on a large amount of collected data, and the preset parameters in the formulas can be set by those skilled in the art according to the actual situation.

[0117] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, ATA hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0118] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0119] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0121] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0122] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0123] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable ATA hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A remote real-time control system for the Industrial Internet, characterized in that, Includes the following modules: The adaptive network delay compensation module is used to monitor network delay through high-frequency heartbeat packets and predict delay trends using an adaptive filtering algorithm; control commands are sent in advance based on the prediction results, and a closed-loop feedback mechanism is introduced to dynamically correct the prediction model. The multi-path redundancy transmission module is used to establish and evaluate the performance of multiple transmission paths. By dynamically calculating the transmission cost, it allocates data packets to the optimal path and automatically switches paths and retransmits when transmission fails, thereby achieving load balancing and reliable transmission. The process by which the multi-path redundant transmission module performs dynamic data packet allocation and load balancing is as follows: Continuously monitor key performance indicators for each transmission path, including: Available bandwidth Current load Current delay , Indicates the first Path; For each data packet to be sent, calculate the sending cost on each path. The formula is as follows: ,in, It is a dynamic load factor used to balance the weights of load and latency. Select sending cost The shortest path is used to send the current data packet; when the cost difference between multiple paths is less than a preset threshold, a round-robin strategy is used to distribute data packets among these high-quality paths to avoid overloading a single path. Dynamically adjust load factor based on network congestion level. : Automatically increase [the number of devices] when a network average latency increase of more than 20% or a packet loss rate increase is detected. A value of λ+0.1 makes the system focus more on load balancing; when network conditions meet the criteria, the value is reduced. The value is λ-0.05, which makes the system pay more attention to transmission delay; The cost of all paths is reassessed every 100ms, and the load on a particular path is detected. When the load exceeds 50% of the average load of other paths, some of the data packets to be sent will be automatically redistributed to the less loaded paths. The intelligent data compression and optimization module is used to select the appropriate compression encoding method according to the characteristics of data types; it dynamically searches for the best compression parameters through a multi-objective optimization function to achieve a balance between compression ratio, processing time and data quality, and adopts differentiated compression strategies for data with different priorities. The intelligent data compression and optimization module performs the optimization mechanism that balances compression quality and efficiency as follows: Real-time monitoring of three key performance indicators: Compression ratio ; Processing time This refers to the time from the start of compression to completion. Data quality ; Construct a comprehensive optimization objective function: ,in: It is a weighted index. To comprehensively optimize the objective function value; Parameter space search and optimization: The compression parameter space is defined as follows: Compression level l; Dictionary size d; Quantization step size q; The optimal solution is searched in the parameter space using the simulated annealing algorithm: initial temperature The cooling coefficient c = 0.95; Perform a neighborhood search at each temperature and accept the probability of a suboptimal solution. ,in The probability of accepting an inferior solution, The change in the objective function value is represented by T, where T is the current temperature in the simulated annealing algorithm; when the temperature drops to... Or stop when the maximum number of iterations is reached; Dynamic weight adjustment mechanism: Adjust the weighting index according to network conditions and application requirements: When network bandwidth meets the requirements: Add ; When network latency exceeds the limit: increase ; Preset key control commands: Add ; Weighting adjustment formula: ,in To adjust the step size, This is the current bandwidth. This is the bandwidth threshold; Continuously monitor the actual compression effect. If the compression performance drops by more than 10%, re-trigger the optimization search, record the historical best parameter combination, and establish a scenario-based parameter library.

2. A remote real-time control system for the Industrial Internet according to claim 1, characterized in that, The execution process of the adaptive network latency compensation module is as follows: First, time synchronization units are deployed at both the control end and the device end to calculate the current network latency value through high-frequency heartbeat packet exchange and record historical latency data to form a latency sequence. Secondly, an adaptive filtering algorithm is used to smooth the delayed sequence, remove noise and jitter, and predict the delay trend of the next time period. Then, the sending time of the control command is adjusted according to the predicted delay, and the command is sent in advance to offset the effect of the delay. At the same time, a dynamic buffer is set to store the command and release it according to the actual arrival time. A closed-loop feedback control mechanism is introduced, which uses the error between the predicted value and the actual measured value as a continuous feedback signal to correct the prediction model itself. Finally, monitor latency changes in real time, and if the latency exceeds the threshold, trigger an alarm and switch to a backup control strategy.

3. A remote real-time control system for the Industrial Internet according to claim 2, characterized in that, The closed-loop feedback control mechanism includes: Error calculation: In each control cycle, when the latest actual network latency measurement is obtained via heartbeat packets... Then, calculate the corresponding predicted value. The error between them ,in, This is the prediction error at the current moment; Feedback control quantity calculation: A digital PI (proportional-integral) controller is used to process this error signal; the PI controller generates the model correction quantity based on the current error (proportional term) and the cumulative error history (integral term); Scale term P: Provides the current error A proportional response, designed to reduce errors immediately; Integral term I: Cumulative past errors , used to eliminate steady-state errors; Control quantity The calculation formula is: in, For production volume, This is the cumulative sum of errors from time 0 to t. It is proportional gain. Integral gain; Online calibration of model parameters: The calculated control inputs are then calibrated online. It applies to the existing adaptive filtering prediction model; When the weight vector of the original adaptive filter is The update rule can then be revised as follows: in, Let be the value of the weight vector of the adaptive filter at time t. It is the step size of the original filter. It is the input vector. It is a mixing factor less than 1; Alternatively, directly correct the predicted output: ,in These are the new predicted values ​​after feedback calibration. These are the original predicted values; Monitoring and Reset: Continuously monitor errors The absolute value or moving average; when the error continues to exceed the preset threshold for a preset time, it is determined that the network has undergone drastic and fundamental changes, and the prediction model is deemed inapplicable at this time; The automatic reset mechanism clears the error accumulation term of the PI controller and restores the filter weights to their default values, allowing the model to start learning and adapting anew on a new basis.

4. A remote real-time control system for the Industrial Internet according to claim 1, characterized in that, The execution process of the multi-path redundant transmission module is as follows: Establish multiple network connections at the control and device ends, including wired, wireless, and mobile network paths, and evaluate the bandwidth, latency, and packet loss rate of each path; The control commands are divided into multiple data packets, and a sequence number and checksum are added to each data packet to perform dynamic data packet allocation and load balancing. The device receives data packets, reassembles instructions according to the sequence number, and verifies data integrity using a checksum. If a transmission path fails or the delay exceeds a threshold, it automatically switches to another path and retransmits the lost data packets. Regularly optimize path selection strategies and dynamically adjust transmission paths based on real-time network performance data to ensure the lowest latency and highest reliability.

5. A remote real-time control system for the Industrial Internet according to claim 1, characterized in that, The execution process of the intelligent data compression and optimization module is as follows: Analyze the types and characteristics of control data, and use differential coding compression for periodic data and dictionary coding compression for non-periodic data; Machine learning algorithms are used to identify redundant patterns in the data and dynamically adjust the compression level to maximize the compression ratio while ensuring data integrity; a balance optimization mechanism between compression quality and efficiency is introduced. Priority tags are added to compressed data before transmission. For example, emergency control instructions use low-latency compression algorithms, while historical logs use high-compression-ratio algorithms. The receiving end decompresses the data and verifies its consistency. If decompression fails, it requests retransmission and adaptively adjusts the compression strategy based on network conditions.

Images