Communication link dual-mode switching control system based on cloud edge collaboration and broadcast subscription

Abstract

The application discloses a communication link dual-mode switching control system based on cloud edge cooperation and broadcast subscription, belongs to the field of engineering machinery intelligent control and wireless communication network cross technology, and comprises the following steps: analyzing matching abnormal index of edge node receiving information package, determining matching abnormal execution strategy; based on multi-screen cooperation management architecture, video stream intelligent distribution and display control are carried out; through hook intelligent tracking technology, real-time visual positioning and automatic zoom control are realized; and a cooperative processing mechanism of cloud end instruction and local execution is executed. The communication link dual-mode switching control system based on cloud edge cooperation and broadcast subscription provided by the application realizes accurate distribution of computing power and bandwidth resources, realizes cross-screen video stream intelligent scheduling and state synchronization through the multi-screen video cooperation management architecture, provides accurate visual positioning and adaptive zoom control based on the hook intelligent tracking technology, and ensures efficient landing of intelligent decision-making by relying on the cooperative mechanism of cloud end instruction and local execution.

Description

Technical Field

[0001] This invention relates to the field of wireless communication network technology, and in particular to a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription. Background Technology

[0002] The existing tower crane cloud-edge collaborative system consists of edge computing nodes and a cloud platform. The edge side (such as the industrial control computer on the tower crane itself) is responsible for real-time data processing, including tower crane pose perception (collecting data through height encoders, nine-axis sensors, tilt sensors, etc.), LiDAR collision avoidance scanning (deploying six radars on the boom to monitor collision risks in real time), and low-latency tasks such as hook video stream analysis. The cloud (industrial center cloud data platform) performs global data fusion and model optimization, such as training tower crane life prediction models, updating collision avoidance algorithms, and remotely controlling edge nodes by issuing commands.

[0003] For example, Chinese invention patent with publication number CN119485414A discloses a hot standby method for carrier and wireless dual-mode communication links, which includes: detecting the current communication quality of network terminal equipment, making a prediction based on the detected historical communication quality of the terminal, adjusting the allocation of different network communication channels according to the predicted communication quality of the terminal at time t+1, and improving the communication quality of the terminal after adjustment.

[0004] For example, Chinese invention patent CN115551116B discloses a method and system for establishing a communication link based on a dual-mode adaptive frequency band, comprising: configuring the frequency band range of a first digital filter of an HPLC to cover a preset communication frequency band; transmitting a signal to a first receiver through the first digital filter of the HPLC, processing the received signal through the first receiver to determine the frequency band in which the signal is located, and determining the first parameters of the first digital filter of the HPLC based on the frequency band; determining the wireless communication frequency band based on the frame protocol of the HPLC; configuring the frequency band range of a second digital filter of an HRF to cover a preset frequency band range; transmitting a signal to a second receiver through the second digital filter of the HRF, processing the received signal through the second receiver to determine the frequency band in which the signal is located, and determining the second parameters of the second digital filter of the HRF based on the frequency band; and establishing a dual-mode communication link based on the first parameters of the first digital filter, the wireless communication frequency band, and the second parameters of the second digital filter.

[0005] The above-mentioned technology has at least the following technical problems:

[0006] When edge computing nodes in existing technologies adopt batch aggregation strategies, critical information is delayed in being sent because it happens to fall at the end of the aggregation cycle. Abnormal information from edge nodes cannot be transmitted to the cloud quickly and reliably, which exacerbates the information silo effect. This makes it difficult for the cloud to obtain global abnormal situations in a timely manner and generate targeted adjustment strategies. Ultimately, the balance between real-time response requirements and global collaborative goals is broken, which can easily lead to delayed decision-making or inappropriate responses. Summary of the Invention

[0007] This invention provides a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription, including: a matching anomaly detection module, used to analyze the matching anomaly indicators of information packets received by edge nodes when monitoring the real-time operation information status of devices, thereby determining the matching anomaly execution strategy;

[0008] The communication link configuration module is used to record the information packets received by the edge node as abnormal information packets when the matching abnormal execution strategy is cloud-edge collaboration, analyze the key indicators of the abnormal information packets, and thus determine the broadcast transmission strategy.

[0009] The communication link adjustment module is used to set broadcast transmission rules when the broadcast transmission strategy is preemptive, adjust the transmission buffer window in combination with key indicators of abnormal information packets, and perform communication link restriction adjustment after transmission.

[0010] The cloud receiving module is used to generate targeted strategies for abnormal information after receiving abnormal information in the cloud.

[0011] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0012] 1. This invention enables edge nodes to have autonomous self-healing capabilities at the anomaly handling level. It can quickly respond to deep matching anomalies by restarting sensor modules and repeatedly analyzing and matching anomaly indicators, and adaptively switching modes, significantly reducing reliance on manual intervention and improving the autonomy and timeliness of anomaly response. In terms of resource scheduling, the local system retains the right to collect core information and releases computing power for caching critical information, while the cloud synchronously prepares dedicated channels, achieving precise allocation of computing power and bandwidth resources. This avoids non-critical information occupying core resources and ensures efficient system operation. In the information transmission stage, by combining the classification of information criticality with a dual-buffer architecture and a preemptive transmission strategy, it ensures that critical information is transmitted preferentially through independent channels, while leveraging flexible windows to adjust and delay the aggregation of non-critical information, optimizing bandwidth utilization while ensuring real-time performance. The cloud achieves accurate diagnosis through multi-dimensional matching.

[0013] 2. This invention establishes a high-priority transmission channel for core control commands by introducing a dual-mode switching and preemptive broadcast mechanism. This mechanism ensures that critical commands can preferentially preempt communication resources and be delivered in a timely manner in complex network environments, effectively avoiding command loss and delays caused by channel congestion or contention, greatly improving the communication success rate, and laying a solid communication foundation for the stable and reliable operation of the entire system.

[0014] 3. This invention, through the deep integration of a multi-screen collaborative management architecture and intelligent hook tracking technology, greatly simplifies the operator's monitoring and control burden. Operators no longer need to perform tedious manual switching and judgment between multiple screens and data sources; the system can automatically allocate video sources, intelligently track the hook, and predict its movement trajectory. This not only significantly improves the accuracy and efficiency of operations but also fundamentally enhances the safety of the operation process through automated monitoring and early warning.

[0015] 4. This invention achieves seamless integration and close collaboration between cloud-based intelligent decision-making and rapid local execution. The cloud handles complex global optimization calculations and efficiently pushes the resulting instructions down to the local system; the local system focuses on rapid response to instructions, security verification, and operational condition adaptation. This collaborative model perfectly combines global optimization with real-time local response, enabling the system to possess both the wisdom of macro-level decision-making and the agility of micro-level execution, resulting in comprehensive optimization of overall operating efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in an embodiment of the present invention;

[0018] Figure 2 This is a method framework diagram of a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in an embodiment of the present invention;

[0019] Figure 3 This is a mind map of the dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in this embodiment of the invention;

[0020] Figure 4 This is a communication network topology diagram of a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in this application embodiment. Detailed Implementation

[0021] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0022] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0023] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] This invention provides a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription, such as... Figure 1 The diagram shows the structure of a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription. The system includes: a matching anomaly detection module, a communication link configuration module, a communication link adjustment module, and a cloud receiving module.

[0027] The matching anomaly detection module is used to analyze the matching anomaly indicators of the information packets received by the edge nodes when monitoring the real-time operation information status of the equipment, thereby determining the matching anomaly execution strategy.

[0028] It should be noted that the information packets received by the edge nodes can be specifically divided into three categories: structural status, operating parameters, and environmental interaction. Structural status information is the key comparison basis, including real-time angle data of the boom's verticality, vibration frequencies of each connection node in the tower body, instantaneous tension values ​​of the wire rope, and wear detection data. Operating parameter information is used to assess the stability of equipment operation, covering the lifting / lowering speed of the hook, the angular velocity of the rotating mechanism, and the pressure values ​​of the hydraulic system. Environmental interaction information relates to the adaptability of the tower crane to the operating environment, including the load weight below the hook, the distance to obstacles within the operating radius, and real-time wind speed.

[0029] Furthermore, the analysis of anomaly indicators in the received packets at the edge nodes is conducted using the following specific analysis methods:

[0030] Obtain the received information from the edge nodes, perform matching analysis, and obtain the matching anomaly parameters.

[0031] It should be noted that the edge nodes extract features from the received sensor information at fixed intervals and calculate the matching degree with the standard templates in the built-in feature library using the cosine similarity algorithm.

[0032] It's worth noting that cosine similarity is an algorithm used to measure the similarity between two vectors. Its core idea is to determine similarity by calculating the cosine of the angle between the two vectors. The closer the cosine value is to 1, the closer the angle between the two vectors is to zero degrees, meaning the two vectors are more similar; conversely, the closer the cosine value is to 0, the lower the similarity.

[0033] Extract the matching similarity threshold. If the cosine similarity is greater than or equal to the matching similarity threshold, it means that the information has a high similarity to the standard template in the built-in feature library, and the tower crane cab is running smoothly. This is recorded as a normal match. If the cosine similarity is less than the matching similarity threshold, it means that the information has a low similarity to the standard template in the built-in feature library. At this time, there is a problem with the operation of the tower crane cab, and precise instructions need to be issued from the cloud.

[0034] Matching abnormal parameters, including the number of non-matching messages, the average non-matching deviation value, and the proportion of non-matching duration.

[0035] It should be noted that the number of mismatches can be calculated by counting the total number of mismatches with a cosine similarity less than the matching similarity threshold within the comparison period using a counter. The average mismatch deviation value can be calculated by subtracting the target value from the actual value when a mismatch is detected, collecting and statistically analyzing the single deviation values ​​of all mismatches within the comparison period, and calculating the arithmetic mean. The mismatch duration percentage can be calculated by starting a timer when a mismatch is first detected, continuously monitoring the matching status, stopping the timer if a successful match is subsequently detected, and calculating the difference between the end time and the start time as the duration of the current consecutive mismatch. If no match is restored by the end of the comparison period, the duration is the difference between the end time of the comparison period and the start time of the first mismatch. The ratio of this duration of consecutive mismatches to the period is then used to obtain the mismatch duration percentage.

[0036] It should be noted that if multiple mismatched durations are connected to matching durations and then to mismatched durations within a period, the duration of each consecutive mismatch is calculated separately, and the total mismatch duration is obtained by adding them together. The ratio of the total mismatch duration to the period is then used to obtain the proportion of mismatch duration.

[0037] It should be noted that the more times a mismatch is detected within the comparison period, the longer the mismatch duration and the larger the proportion of mismatch duration. A higher number of mismatches may also mean multiple consecutive mismatches, and the total number of mismatches will increase due to multiple comparisons, resulting in a longer cumulative duration. The average mismatch deviation value may be positively correlated with the number of mismatches. A higher number of mismatches is accompanied by a larger average deviation value. The larger the deviation value, the more difficult it is for the system to automatically correct the mismatch, which in turn leads to an increase in the duration of continuous mismatches.

[0038] It should be noted that a higher number of mismatches means a higher frequency of matching failures within the statistical period, reflecting a poorer stability of the tower crane and a significant increase in the matching anomaly index. A larger average mismatch deviation value indicates a more significant difference in the data at each mismatch, indicating a higher severity of the mismatch and driving up the matching anomaly index. A larger proportion of mismatch duration means a longer duration of mismatch, a weaker self-healing ability of the tower crane system, and an inability to quickly restore consistency, causing the matching anomaly index to rise due to the increased persistence of the anomaly.

[0039] The matching anomaly indicators of the edge node's received information are analyzed based on the matching anomaly parameters.

[0040] Extract the influence factors of the number of mismatches, the average mismatch deviation value, and the proportion of mismatch duration from the database.

[0041] It should be noted that the values ​​of the influence factors for the frequency of mismatched information, the average mismatch deviation value, and the proportion of mismatch duration are all between 0 and 1, and the sum of these three factors is 1. When using these factors, pre-defined values ​​can be directly extracted from the database. For example, a one-to-one mapping set can be constructed between the frequency of mismatched information, the average mismatch deviation value, and the proportion of mismatch duration, and their corresponding influence factors. When using these factors, the real-time acquired frequency of mismatched information, the average mismatch deviation value, and the proportion of mismatch duration are input into the corresponding mapping set, thereby extracting the influence factors for the frequency of mismatched information, the average mismatch deviation value, and the proportion of mismatch duration.

[0042] Extract the reference values ​​for the number of mismatches, the average mismatch deviation, and the percentage of mismatch duration from the database.

[0043] The matching anomaly index of edge node received information is a quantitative indicator that measures the impact of the number of mismatches, the average mismatch deviation, and the proportion of mismatch duration on the matching anomaly index. The specific analysis process is as follows: The number of mismatches, the average mismatch deviation, and the proportion of mismatch duration are compared with the corresponding reference values. The results of each comparison are then coupled with the corresponding feature influencing factors to obtain the matching anomaly index of edge node received information.

[0044] ;

[0045] Where F represents the matching anomaly index of the information received by the edge node, Z represents the number of mismatches, Z0 represents the reference value of the number of mismatches, a represents the characteristic influence factor of the number of mismatches, P represents the average mismatch deviation value, P0 represents the reference value of the average mismatch deviation value, s represents the characteristic influence factor of the average mismatch deviation value, C represents the proportion of mismatch duration, C0 represents the reference value of the proportion of mismatch duration, and d represents the characteristic influence factor of the proportion of mismatch duration.

[0046] Furthermore, the matching exception execution strategy is determined accordingly, and the specific analysis method is as follows:

[0047] Extract the preset threshold values ​​for matching anomalies from the database.

[0048] If the matching anomaly indicator is greater than or equal to the threshold, the matching anomaly execution strategy will be recorded as cloud-edge collaborative adjustment.

[0049] It should be noted that if the matching anomaly index is greater than or equal to the threshold, it indicates that the matching anomaly of the information packet is relatively deep, which means that there is a fundamental and complex problem with the data consistency. At this time, relying solely on local adjustment is insufficient to address the core of the problem and cannot completely solve it. It is necessary to enter cloud-edge collaborative adjustment to conduct multi-dimensional source tracing analysis of the anomaly. Therefore, the matching anomaly execution strategy is referred to as cloud-edge collaborative adjustment.

[0050] If the matched abnormal indicator is less than the threshold, the preset abnormal threat range in the database is extracted. The upper limit of the range is the threshold of the matched abnormal indicator. If the matched abnormal indicator is less than the lower limit of the abnormal threat range, the matched abnormal execution strategy is recorded as no mode adjustment is required.

[0051] It should be noted that if the matching anomaly index is less than the lower limit of the anomaly threat range, it means that the degree to which the information packet deviates from the normal matching state is extremely limited. This kind of anomaly usually falls within the acceptable error range that is difficult to completely avoid during normal system operation. Within the tolerance of the local feature library, there is no need to waste computing resources to adjust it. Therefore, the matching anomaly execution strategy is recorded as no mode adjustment required.

[0052] If the matched abnormal indicator is within the abnormal threat range, the matched abnormal execution strategy will be recorded as executing sensor restart.

[0053] It should be noted that if the anomaly indicator is within the abnormal threat range, it means that the information packet deviates significantly from the normal matching state. The anomaly indicator is close to the anomaly indicator threshold, which means that if left unchecked, it is highly likely to jump to an anomaly state in a short period of time, at which point local adjustments alone will be completely ineffective. The danger of this critical state lies in its uncertainty, so the anomaly execution strategy is recorded as executing a sensor restart.

[0054] Furthermore, the sensor restart strategy is implemented, and the specific analysis method is as follows:

[0055] The number of re-analysis attempts is extracted based on the matched abnormal indicators.

[0056] It should be noted that the number of analyses corresponding to each interval of the matched abnormal indicators is extracted from the database, and the number of analyses corresponding to the intervals where the matched abnormal indicators are located is also extracted and named as the number of re-analysis.

[0057] It should be added that the larger the abnormal matching index, the closer it is to the abnormal matching index threshold, the greater the deviation from the normal matching state, and the greater the degree of vigilance required. In order to ensure accurate diagnosis after restarting, the corresponding number of extractions will be more.

[0058] A sensor restart signal is generated. After waiting for confirmation, the matching anomaly indicators are re-analyzed based on the number of re-analysis attempts to obtain a set of matching anomaly indicators. The maximum value of the matching anomaly indicator is then extracted. If it is still within the anomaly threat range, it is upgraded to a matching anomaly. The matching anomaly execution strategy is recorded as cloud-edge collaborative adjustment.

[0059] It should be noted that when a sensor restart signal is generated, the cockpit is prompted to confirm the sensor restart. After waiting for the confirmation information, the matching anomaly indicators are re-analyzed based on the number of re-analysis attempts, resulting in multiple matching anomaly indicators. A set of matching anomaly indicators is constructed, and the maximum value of the anomaly indicator is extracted from the set. If the maximum value of the anomaly indicator is still within the anomaly threat range, it indicates that the information packet deviates significantly from the normal matching state after the sensor restart. If the matching anomaly indicator is close to the matching anomaly indicator threshold, it indicates that the cause of the information packet deviation is not sensor fluctuation. Therefore, it is upgraded to a matching anomaly, and the matching anomaly execution strategy is recorded as cloud-edge collaborative adjustment.

[0060] If the value is less than the lower limit of the abnormal threat range, the matching abnormal execution strategy will be recorded as no mode adjustment is required, and a critical prompt message will be generated.

[0061] It should be noted that if the value is less than the lower limit of the abnormal threat range, it means that the matching anomaly of the information packet has been improved after restarting the sensor, and the matching effect is good, which confirms the effectiveness of the measure of restarting the sensor. Therefore, the matching anomaly execution strategy is recorded as no mode adjustment is required, and the critical prompt information of the information packet is generated.

[0062] In this embodiment, the critical warning message could be "Note that this data was close to matching the abnormal indicator threshold".

[0063] It should be noted that the system has a cockpit-based direct network connection operation mode and a cloud-based intelligent driving mode, and can seamlessly switch between control consoles according to network conditions, application scenarios, task requirements, or command priorities to ensure control continuity and reliability.

[0064] This invention achieves refined differentiation of matching anomaly states by setting dual judgment criteria of thresholds and anomaly threat ranges, avoiding simplistic black-and-white judgments and enabling more accurate identification of the degree of anomaly, providing a clear basis for subsequent processing. Secondly, for situations within the anomaly threat range, a sensor module restart self-healing mechanism is introduced, giving the system autonomous repair capabilities, reducing reliance on external intervention, improving the efficiency of anomaly handling, and curbing the trend of anomaly escalation in the early stages. Furthermore, after restarting, by analyzing matching anomaly indicators multiple times and taking the maximum value, the risk of misjudgment is further reduced, ensuring the rigor of the decision to upgrade to matching anomaly, and providing a reliable trigger basis for subsequent higher-level collaborative processing. Simultaneously, when the indicator is less than the lower limit of the range, it is recorded as normal and a critical warning is generated, clearly defining the normal state while also alerting to potential risks, achieving effective monitoring of critical states, balancing system stability and risk warning, and making the overall anomaly handling process more hierarchical and reliable.

[0065] The communication link configuration module is used to record the information packets received by the edge node as abnormal information packets when the matching abnormal execution strategy is cloud-edge collaboration, analyze the key indicators of the abnormal information packets, and thus determine the broadcast transmission strategy.

[0066] Furthermore, the key indicators of the abnormal information packets were analyzed, and the specific analysis methods are as follows:

[0067] Obtain key data from the anomaly information packet, including message type importance, real-time sensitivity, and timestamp freshness.

[0068] It should be noted that message type importance refers to the degree of criticality of a message in the system. By identifying the message type label, matching it with a predefined importance level system in the database, and automatically assigning the corresponding quantitative value.

[0069] Real-time sensitivity refers to the tolerance of message processing for latency. The system analyzes the business attributes of messages, combines the information value loss data caused by latency in historical interactions, collects the correspondence between latency time and information value loss data in historical data, solves the linear function using the least squares method, and establishes a sensitivity evaluation model. The real-time requirements under different scenarios are transformed into quantifiable sensitivity indicators. The higher the indicator, the lower the tolerance for latency and the higher the real-time sensitivity.

[0070] It should be noted that the expression for a linear function is: , where p represents the information value loss rate after the delay time, k represents the sensitivity coefficient, x represents the delay time, b represents the initial loss, and k and b are used to minimize the sum of the squared differences between the model prediction value and the actual value of historical data by the least squares method to solve for the optimal parameters.

[0071] Timestamp freshness refers to the interval between the time the data is generated and the time it is processed. The system embeds a high-precision timestamp at the message generation node and calculates the difference between the timestamp and the current receiving time by comparing it with the local standard time at the receiving end. The smaller the difference, the higher the freshness. At the same time, the system further calibrates the quantitative value of freshness by combining the effective window period parameter of the message to ensure that it can truly reflect the timeliness of the information.

[0072] It should be noted that the greater the importance of a message, the more stringent the real-time requirements become. This stringency is not only reflected in the extreme pursuit of transmission speed and the precision of end-to-end timeliness control, where even the slightest delay can devalue high-value information, but also in the stronger the dependence of real-time on freshness. As the core carrier of freshness, the accuracy of the timestamp is elevated to a decisive position. At the same time, the effective boundaries of real-time are also clearer, and must be strictly confined within the freshness range defined by the timestamp. Otherwise, no matter how fast the transmission speed, it cannot give the message any real meaning.

[0073] It's important to note that the greater the importance of the message type, the higher the critical indicators of the abnormal information packet. Highly important messages are often associated with core decisions, significant interests, or urgent actions; their abnormal states have a more profound impact on the system, thus their corresponding critical indicators will increase, becoming one of the core factors in measuring the importance of the information. The greater the real-time sensitivity, the higher the critical indicators of the abnormal information packet. High real-time sensitivity means that the information has extremely low tolerance for delays. Once an anomaly occurs, if it is not handled promptly, the information's value will rapidly diminish or even become invalid, potentially leading to serious consequences. Therefore, its critical indicators will significantly increase due to the stringent requirements for timeliness. The greater the timestamp freshness, the higher the critical indicators of the abnormal information packet. High timestamp freshness indicates that the difference between the information's generation time and the current time is small, falling within the effective window period. At this time, the abnormal state can more accurately reflect the current system's true situation, providing stronger reference value for real-time decision-making, thereby driving up the critical indicators.

[0074] Analyze the key indicators of the anomaly information packets based on the data.

[0075] Extract the message type importance feature influence factor, real-time sensitivity feature influence factor, and timestamp freshness feature influence factor from the database.

[0076] It should be noted that the values ​​of the message type importance feature influence factor, real-time sensitivity feature influence factor, and timestamp freshness feature influence factor all range from 0 to 1, and the sum of these three features is 1. When using these features, pre-defined values ​​can be directly extracted from the database. For example, a one-to-one mapping set can be constructed between message type importance, real-time sensitivity, and timestamp freshness and their corresponding message type importance feature influence factors, real-time sensitivity feature influence factors, and timestamp freshness feature influence factors. When using these features, the real-time message type importance, real-time sensitivity, and timestamp freshness values ​​are input into the corresponding mapping sets to extract the message type importance feature influence factors, real-time sensitivity feature influence factors, and timestamp freshness feature influence factors.

[0077] Extract preset reference values ​​for message type importance, real-time sensitivity, and timestamp freshness from the database.

[0078] The key indicators of abnormal information packets are quantitative indicators of the degree of influence of message type importance, real-time sensitivity and timestamp freshness on the key indicators of abnormal information packets. The specific analysis process is as follows: The collected message type importance, real-time sensitivity and timestamp freshness are compared with the corresponding reference values. The results of each processing are coupled with the corresponding feature influence factors to obtain the key indicators of abnormal information packets.

[0079] ;

[0080] Wherein, G represents the key indicator of the abnormal information packet, Q represents the importance of the message type, Q0 represents the reference value of the message type importance, r represents the influence factor of the message type importance feature, W represents the real-time sensitivity, W0 represents the reference value of the real-time sensitivity feature, t represents the influence factor of the real-time sensitivity feature, H represents the timestamp freshness, H0 represents the reference value of the timestamp freshness feature, and y represents the influence factor of the timestamp freshness feature.

[0081] Furthermore, the broadcast transmission strategy is determined accordingly, and the specific analysis method is as follows:

[0082] Extract the key indicator thresholds of the preset abnormal information packets from the database.

[0083] If the critical indicator of the abnormal information packet is greater than or equal to the critical indicator threshold of the abnormal information packet, the abnormal information packet is recorded as a critical information packet, and the broadcast transmission strategy is recorded as preemptive transmission.

[0084] It should be noted that if the critical indicators of the abnormal information packet are greater than or equal to the critical indicator threshold of the abnormal information packet, it indicates that this abnormal information has extremely critical attributes. If transmission delay or packet loss occurs, it may cause the fault propagation coefficient to increase exponentially. Therefore, it is necessary to send the information to the cloud quickly, which is called the preemptive transmission strategy.

[0085] If the criticality index of an abnormal information packet is less than the criticality index threshold of an abnormal information packet, the abnormal information packet is recorded as a non-critical information packet, and the broadcast transmission strategy is recorded as normal transmission.

[0086] It should be noted that if the criticality index of the abnormal information packet is less than the criticality index threshold of the abnormal information packet, it indicates that the abnormal information is non-critical information. Such information usually belongs to occasional disturbance messages during system operation. It does not require the use of emergency transmission resources and can be included in the regular transmission queue to wait for scheduling. It can be processed during the interval of critical information transmission or during periods of resource idleness. Its transmission priority is automatically lower than that of critical information. Therefore, the broadcast transmission strategy is recorded as normal transmission.

[0087] The communication link adjustment module is used to set broadcast transmission rules when the broadcast transmission strategy is preemptive, adjust the transmission buffer window in combination with key indicators of abnormal information packets, and perform communication link restriction adjustment after transmission.

[0088] Furthermore, broadcast transmission rules are set, and the specific analysis method is as follows:

[0089] It should be noted that a dual-buffer architecture is used to distinguish information types. One buffer is used to cache non-critical information to be aggregated and is transmitted to the cloud through the ordinary link of the broadcast link; the other dedicated buffer handles critical information and is directly connected to the independent MQ broadcast link of the broadcast link, using a "preemptive sending" strategy.

[0090] If the broadcast transmission strategy is preemptive, the broadcast transmission rule is set to direct connection to an independent broadcast link. The current link status is checked, and if there is information transmission on the current broadcast link, it waits. During the waiting period, the same level information that arrives is merged and sent to the cloud through the broadcast link.

[0091] New critical information packets arriving during transmission are immediately broadcast to the cloud via a broadcast link after the transmission ends.

[0092] It should be noted that the preemptive sending strategy waits if there is information being transmitted on the current broadcast link. During the waiting period, information of the same priority that arrives is merged and sent to the cloud through the broadcast link. New critical information that arrives during the transmission process is sent to the cloud through the broadcast link immediately after the current transmission ends.

[0093] If the broadcast transmission strategy is normal transmission, then the broadcast transmission rule is set to normal link of broadcast link. After preemptive transmission, the broadcast protocol enters the confirmation waiting phase. At this time, the edge node immediately activates the lagging aggregation engine.

[0094] It should be noted that after preemptive sending, the MQ broadcast protocol enters the confirmation waiting phase. At this time, the edge node immediately activates the lagging aggregation engine to compress the accumulated non-critical information.

[0095] The length of the aggregation time period is determined based on the key indicators of the anomaly information packet.

[0096] It should be noted that the time period lengths corresponding to the key indicator intervals of each anomaly information package are extracted from the database, and the time period lengths corresponding to the intervals where the key indicators of the anomaly information package are located are mapped and named as the aggregated time period length.

[0097] It should be added that the larger the critical indicators of the abnormal information packet, the higher its importance to the system, and the greater the attention paid to it. In order to ensure the accuracy of the system, the corresponding aggregation time period should be shorter.

[0098] The non-critical information packet set is obtained by aggregating the time length, and the non-critical information generated within the time length is aggregated and compressed. The critical information packets are then sent to the cloud during idle time slots after being transmitted through the broadcast link.

[0099] It should be noted that after the critical information is triggered for immediate transmission, the non-critical information accumulated over the subsequent aggregation time is compressed into a single frame and sent to the cloud using the idle time slot after the critical information is transmitted through the broadcast link. This ensures the timeliness of the critical information while reducing additional bandwidth overhead.

[0100] This invention employs a dual-buffer architecture to clearly distinguish information types, allowing non-critical information to be transmitted in an orderly manner through ordinary channels, while critical information is processed exclusively through a dedicated MQ broadcast link. This fundamentally avoids transmission conflicts between information of different priorities, ensuring that critical information is not overwhelmed by non-critical information, thus laying the foundation for efficient, layered information transmission. The application of a preemptive sending strategy further strengthens the timeliness guarantee of critical information: the waiting mechanism when the current channel is busy merges with the transmission of information of the same priority, avoiding channel congestion and reducing the number of transmissions; while the rule of sending new critical information immediately after the current transmission ends minimizes the transmission delay of high-priority information, ensuring that its core value is not lost.

[0101] Furthermore, the transmission buffer window is adjusted, and the specific analysis method is as follows:

[0102] When there are no critical information packets, maintain a normal window to optimize bandwidth utilization efficiency.

[0103] It should be noted that when there are no critical information packets and low-priority small data packets are transmitted sporadically, the protocol header overhead of each frame is too high, which will lead to a significant decrease in bandwidth utilization. At the same time, frequent frame exchanges will increase the signaling overhead of wireless or wired links, wasting valuable transmission resources. More importantly, reserving redundant bandwidth can ensure that subsequent bursts of critical information packets can quickly seize the transmission link and avoid delays caused by low-value data occupying bandwidth. Therefore, information within a fixed window is compressed into a single frame and sent uniformly.

[0104] When a critical information packet is detected, the transmission buffer window is adjusted, and the transmission buffer window value is determined based on the critical indicators of the abnormal information packet.

[0105] It should be noted that the window buffer values ​​corresponding to the key indicator ranges of each abnormal information packet are extracted from the database. At the same time, the window buffer values ​​corresponding to the key indicator ranges of the abnormal information packets are mapped and extracted, and named as transmission buffer window buffer values.

[0106] It should be added that the larger the critical indicator of the abnormal information packet, the more critical the role of the information in the system, and the higher its importance. In order to send the critical information packet quickly, the corresponding window buffer value should be smaller.

[0107] The new transmission buffer window size is obtained based on the window size corresponding to the key information packet at the window position and the transmission buffer buffer value, and the information is mapped to the independent broadcast link.

[0108] It should be noted that the new transmission buffer window size is obtained by adding the transmission buffer buffer value to the window size corresponding to the key information packet at the window position. The information is then mapped to the independent broadcast link and sent to the cloud based on the new transmission buffer window size.

[0109] It should be noted that analyzing key indicators of abnormal packets to quantify their criticality not only helps identify the transmission pattern of abnormal packets, reducing resource consumption and preventing non-critical packets from occupying processing resources meant for critical packets, but also allows for determining the duration of non-critical packets based on these indicators. By dynamically adjusting the aggregation duration of non-critical packets, resource utilization efficiency can be maximized, and resource consumption can be avoided. Furthermore, the window size for transmitting critical packets can be determined based on these indicators, optimizing the precise allocation of network resources and improving the efficiency, real-time performance, and reliability of data transmission.

[0110] Furthermore, communication link limitation adjustments are performed, and the specific analysis method is as follows:

[0111] When edge nodes transmit to the cloud via broadcast links, they extract the preset threshold for the number of broadcasts of key information packets per unit time from the database.

[0112] If the number of broadcasts of critical information packets exceeds the broadcast frequency threshold within a unit of time, a downgrade is triggered. The specific process is as follows: some critical information packets are switched to a batch transmission mode via the broadcast link, and the non-critical aggregation window is dynamically expanded.

[0113] It should be noted that if the number of broadcasts of critical information packets per unit time exceeds the broadcast frequency threshold, it means that the transmission frequency of the current critical information has exceeded the bandwidth capacity threshold, and the bandwidth has entered an unstable fluctuation range. In order to maintain bandwidth stability, degradation processing is initiated, some critical information is compressed and encapsulated, and sent in batches through the broadcast link to reduce the sending frequency. Based on the real-time number of current critical information packets, the corresponding non-critical aggregation window baseline value is extracted, and the window is dynamically widened proportionally. By extending the aggregation period of non-critical information, the carrying capacity of non-critical information in a single frame is increased, thereby reducing the sending frequency of non-critical information.

[0114] If the number of times the critical information packet is broadcast within a unit of time does not exceed the broadcast number threshold, then transmission continues.

[0115] It should be noted that if the number of broadcasts of critical information packets per unit time does not exceed the broadcast frequency threshold, it indicates that the current transmission frequency of critical information is within the safe bandwidth carrying range, the transmission of various types of data on the link is orderly, and the allocation of bandwidth resources is in an optimal and balanced state. Based on this, there is no need to adjust the transmission mode, and continuing to maintain the current transmission mode is the optimal choice.

[0116] The cloud receiving module is used to generate targeted strategies for abnormal information after receiving abnormal information in the cloud.

[0117] Furthermore, a targeted strategy for generating anomaly information is developed, and the specific analysis method is as follows:

[0118] After receiving the abnormal information packet in the cloud, it calls the global feature library to perform multi-dimensional matching.

[0119] If a match is found with the preset sub-database of environmental change features in the database, the cloud generates an anomaly information package with a targeted strategy of adjusting the matching analysis threshold.

[0120] It should be noted that if the comparison with the feature sub-database of drastic environmental changes is successful, it means that the anomaly is not caused by hardware failure or software logic defects in the cockpit system, but by sudden and drastic changes in the external environment, which causes the raw data collected by the sensors to deviate from the normal baseline, thereby triggering a matching anomaly alarm. At this time, the matching threshold is adaptively widened on the original basis so that the threshold range matches the reasonable data fluctuation range in the current environment.

[0121] The device corresponding to the abnormal information packet is identified and denoted as the target device.

[0122] The historical cycle length of the target device is determined based on matching anomaly indicators.

[0123] It should be noted that the historical period lengths corresponding to each matching abnormal indicator interval are extracted from the database, and the historical period lengths corresponding to the intervals where the matching abnormal indicators are located are mapped and named as the target device historical period lengths.

[0124] It should be added that the larger the anomaly index, the greater the deviation of the information from the normal match. In order to accurately identify the specific root cause of the anomaly, the longer the historical period extracted should be.

[0125] It should be noted that analyzing and matching anomaly indicators can not only determine whether the execution strategy is local adjustment or cloud-edge collaboration, but also retrieve data on demand, accurately pinpoint the time range related to anomalies, avoid including periods that are too long or too short, thereby reducing the interference of invalid data on the analysis, and making the selection of historical data more in line with the actual needs of anomaly analysis, thereby improving the accuracy and efficiency of anomaly diagnosis and trend prediction.

[0126] Select the target device's state pattern within a historical period. If the comparison with the target device's historical state pattern is successful...

[0127] It should be noted that if the comparison with the historical state pattern of the target device is successful, it means that the anomaly is caused by a hardware failure or software logic defect in the cockpit system. At this time, the cloud calls the full historical data of the target device in distributed storage, locates the specific root cause of the anomaly through time-series comparison algorithm, and generates precise intervention instructions based on the comparison results.

[0128] The similarity between the abnormal information packet and each historical state pattern is obtained to obtain the abnormal information packet similarity set.

[0129] Extract the preset similarity threshold from the database.

[0130] Traverse the set of similarity scores for abnormal information packets, select the maximum similarity score, and compare the maximum similarity score with the similarity threshold.

[0131] If the maximum similarity is greater than or equal to the similarity threshold, preventative maintenance is activated.

[0132] It should be noted that if the maximum similarity is greater than or equal to the similarity threshold, it means that the matching reliability between the feature vector of the current anomaly and historical failure cases is high, and the specific root cause of the anomaly has a high degree of accuracy. Based on this, the system can activate the preventive maintenance mechanism, and use historical data to deduce the evolution path of the current anomaly, and intervene in advance to prevent the failure from escalating.

[0133] If the maximum similarity is less than the similarity threshold, an early warning message is generated, and the system immediately downgrades to a lower level.

[0134] It should be noted that if the maximum similarity is less than the similarity threshold, it indicates that the feature vector of the current anomaly information matches all known fault features in the historical state pattern library at a low level. This means that the true root cause of the anomaly may not be within the coverage of the historical state patterns, and the existing diagnostic model of the system is unable to accurately predict its evolution path and scope of impact. In this case, relying on historical experience for conventional processing may lead to misjudgment due to matching errors, or even delay the intervention time, allowing the local anomaly to spread into a systemic failure. Based on this, an early warning message is generated, and the operating load is immediately reduced to control the risk boundary.

[0135] In this embodiment, the warning message could be: "Attention! An atypical anomaly has occurred."

[0136] It should be noted that the cockpit is not directly connected to the tower crane itself, but rather obtains key command and status information, such as warnings, from the tower control cloud APP by subscribing to MQ broadcasts. The cloud APP sends commands to the cockpit via MQ broadcasts, and the cockpit, by subscribing to the "broadcast information" of the cloud APP, becomes a local information aggregation point for the entire tower crane group, realizing localized collision avoidance warnings and dynamic scheduling based on a global perspective. This serves as a localized execution and supplement to intelligent cloud-based driving. The cockpit makes deep use of the intelligent driving capabilities (obstacle avoidance, path planning, etc.) already implemented in the cloud APP, using the advanced commands / decision results (such as obstacle avoidance commands and waypoints) issued by the cloud APP as the basis for cockpit execution, while focusing on command parsing, status monitoring, and human-machine interaction optimization.

[0137] like Figure 2 As shown in the flowchart of the method for a dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in this application embodiment, the method includes: firstly, analyzing the matching anomaly indicators of the information packets received by the edge nodes, thereby determining the matching anomaly execution strategy; then, recording the information packets received by the edge nodes as anomaly information packets, analyzing the key indicators of the anomaly information packets, thereby determining the broadcast transmission strategy; when the broadcast transmission strategy is preemptive transmission, setting the broadcast transmission rules; adjusting the transmission buffer window in combination with the key indicators of the anomaly information packets; performing communication link restriction adjustment after transmission; and finally, generating a targeted strategy for the anomaly information after receiving the anomaly information in the cloud.

[0138] like Figure 3As shown in the mind map of the dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription provided in this embodiment of the invention, the system includes: firstly, analyzing the matching anomaly indicators of the information packets received by the edge nodes, thereby determining the matching anomaly execution strategy; when the matching anomaly execution strategy is cloud-edge collaboration, recording the information packets received by the edge nodes as anomaly information packets; analyzing the key indicators of the anomaly information packets, thereby determining the broadcast transmission strategy; then setting the broadcast transmission rules; adjusting the transmission buffer window in combination with the key indicators of the anomaly information packets; performing communication link restriction adjustment after transmission; and finally generating a targeted strategy for anomaly information.

[0139] like Figure 4 The diagram shown is a communication network topology of a dual-mode switching control system based on cloud-edge collaboration and broadcast subscription provided in this application embodiment. It includes a multi-screen collaborative management architecture, intelligent hook tracking technology, and cloud-based collaborative execution mechanism. The multi-screen video collaborative management architecture includes intelligent screen mapping: automatic identification and matching based on screen physical coordinates, supporting plug-and-play functionality for heterogeneous display devices.

[0140] Dynamic video stream allocation: intelligently allocates video sources to different screens based on business scenarios, with the main screen displaying panoramic monitoring and the secondary screen focusing on specific information.

[0141] State synchronization mechanism: Through a unified CockpitStatusData state manager, consistent updates of displayed content are ensured across multiple screens.

[0142] Cross-screen interaction: Supports cross-screen operations such as touch gestures and keyboard shortcuts, enabling collaborative control and information interaction between multiple screens.

[0143] The intelligent hook tracking system includes multi-view visual fusion: integrating multiple video sources such as a panoramic camera of the crane boom, a close-up camera of the hook, and a collision avoidance monitoring camera.

[0144] Real-time spatial positioning: The spatial position of the hook is calculated using the GetAngleDegrees algorithm, and precise positioning is achieved by combining height and amplitude parameters.

[0145] Adaptive zoom control: Based on the hook movement speed and working height, the optical zoom is dynamically adjusted to maintain the best viewing angle.

[0146] Motion trajectory prediction: Analyze historical motion data to predict the future position of the hook, achieving smooth tracking and anti-shake control.

[0147] The collaboration between cloud commands and local execution includes intelligent decision-making: complex decisions such as path planning and collision avoidance calculations are made in the cloud, and the results are broadcast to the cockpit via MQTT.

[0148] Local rapid response: The cockpit executes commands immediately upon receiving them from the cloud, while simultaneously performing safety verification and operational condition adaptation.

[0149] Two-way status synchronization: The local operating status is reported to the cloud in real time, and the cloud optimizes decision-making instructions based on the overall situation.

[0150] Fault tolerance and degradation handling: In the event of a network outage, the local system continues to operate safely based on the last valid instruction and preset rules.

[0151] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), 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 the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0152] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0153] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0154] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers 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 the present invention.

[0155] 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 implementations should not be considered beyond the scope of this invention.

[0156] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0157] In the several embodiments provided by this invention, it should be understood that the disclosed devices, 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 device, 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 devices or units may be electrical, mechanical, or other forms.

[0158] 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; that is, 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 according to actual needs.

[0159] In addition, the functional units in the various embodiments of the present invention 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.

[0160] 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 invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0161] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription, characterized in that, The system includes: The matching anomaly detection module is used to analyze the matching anomaly indicators of the information packets received by the edge nodes when monitoring the real-time operation information status of the equipment, thereby determining the matching anomaly execution strategy. The analysis method for identifying anomalies in the received packets at the edge nodes is as follows: The received information packets from the edge nodes are acquired, and matching analysis is performed to obtain the matching anomaly parameters; The matching anomaly parameters include the number of non-matching messages, the average non-matching deviation value, and the proportion of non-matching duration. Analyze the matching anomaly indicators of the received packets by the edge nodes based on the matching anomaly parameters; The matching anomaly index of the information packets received by the edge node is a quantitative indicator that measures the degree of influence of the number of mismatches, the average mismatch deviation value, and the proportion of mismatch duration on the matching anomaly index. The specific analysis process is as follows: The number of mismatches, the average mismatch deviation value, and the proportion of mismatch duration are compared with the corresponding reference values. The comparison results are then coupled with the corresponding feature influence factors to obtain the matching anomaly index of the information packets received by the edge node. The communication link configuration module is used to record the information packets received by the edge node as abnormal information packets when the matching abnormal execution strategy is cloud-edge collaboration, analyze the key indicators of the abnormal information packets, and thus determine the broadcast transmission strategy. The communication link adjustment module is used to set broadcast transmission rules when the broadcast transmission strategy is preemptive, adjust the transmission buffer window in combination with key indicators of abnormal information packets, and perform communication link restriction adjustment after transmission. The cloud receiving module is used to generate targeted strategies for abnormal information packets after receiving them in the cloud.

2. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 1, characterized in that, The specific analysis method for determining the matching exception execution strategy is as follows: Extract the preset threshold values ​​for matching anomalies from the database; If the matching anomaly indicator is greater than or equal to the threshold, the matching anomaly execution strategy will be recorded as cloud-edge collaborative adjustment; If the matched abnormal indicator is less than the threshold, the preset abnormal threat range in the database is extracted. The upper limit of the range is the matched abnormal indicator threshold. If the matched abnormal indicator is less than the lower limit of the abnormal threat range, the matched abnormal execution strategy is recorded as no mode adjustment is required. If the matched abnormal indicator is within the abnormal threat range, the matched abnormal execution strategy will be recorded as executing sensor restart.

3. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 2, characterized in that, The specific analysis method for restarting the sensor is as follows: The number of re-analysis attempts is extracted based on the matched anomaly indicators; Generate a sensor restart signal, wait for confirmation information, re-analyze the matching anomaly indicators based on the number of re-analysis attempts, obtain the set of matching anomaly indicators, extract the maximum value of the matching anomaly indicators, and if it is still within the anomaly threat range, upgrade it to a matching anomaly and record the matching anomaly execution strategy as cloud-edge collaborative adjustment. If the value is less than the lower limit of the abnormal threat range, the matching abnormal execution strategy will be recorded as no mode adjustment is required, and a critical prompt message will be generated.

4. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 1, characterized in that, The key indicators of the analyzed anomaly information packets are analyzed using the following specific methods: Obtain key data from the abnormal information packet, including message type importance, real-time sensitivity, and timestamp freshness; Analyze the key indicators of the abnormal information packet based on the key data of the abnormal information packet; The key indicators of the abnormal information packet are quantitative indicators of the degree of influence of message type importance, real-time sensitivity and timestamp freshness on the key indicators of the abnormal information packet. The specific analysis process is as follows: the collected message type importance, real-time sensitivity and timestamp freshness are compared with the corresponding reference values, and the processing results are coupled with the corresponding feature influence factors to obtain the key indicators of the abnormal information packet.

5. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 1, characterized in that, The broadcast transmission strategy is thus determined, and the specific analysis method is as follows: Extract the preset threshold values ​​for key indicators from the abnormal information packets in the database; If the critical indicator of the abnormal information packet is greater than or equal to the critical indicator threshold of the abnormal information packet, the abnormal information packet is recorded as a critical information packet, and the broadcast transmission strategy is recorded as preemptive transmission. If the criticality index of an abnormal information packet is less than the criticality index threshold of an abnormal information packet, the abnormal information packet is recorded as a non-critical information packet, and the broadcast transmission strategy is recorded as normal transmission.

6. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 5, characterized in that, The specific analysis method for setting broadcast transmission rules is as follows: If the broadcast transmission strategy is preemptive, the broadcast transmission rule is set to direct connection to an independent broadcast link. The current link status is checked. If there is information transmission on the current broadcast link, the system waits. During the waiting period, the same-level information that arrives is merged and sent to the cloud through the broadcast link. New critical information packets arriving during the transmission process are immediately broadcast to the cloud via a broadcast link after the current transmission is completed; If the broadcast transmission strategy is normal transmission, then the broadcast transmission rule is set to the normal link of the broadcast link. After preemptive transmission, the broadcast protocol enters the confirmation waiting phase. At this time, the edge node immediately activates the lag aggregation engine. The length of the aggregation time period is determined based on the key indicators of the anomaly information packet; Obtain a set of non-critical information packets based on the aggregation time period length, aggregate and compress the non-critical information generated within the aggregation time period length, and send it to the cloud using the idle time slots after the critical information packets are transmitted through the broadcast link.

7. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 6, characterized in that, The specific analysis method for adjusting the transmission buffer window is as follows: When there are no critical information packets, maintain the normal window to optimize bandwidth utilization efficiency; When a critical information packet is detected, the transmission buffer window is adjusted, and the transmission buffer window value is determined based on the critical indicators of the abnormal information packet. Based on the window size corresponding to the critical information packet's position in the window and the transmission buffer buffer value, a new transmission buffer window size is obtained, and the critical information packet is mapped to an independent broadcast link.

8. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 7, characterized in that, The specific analysis method for adjusting the communication link limitation is as follows: When edge nodes transmit to the cloud via broadcast links, the threshold number of broadcasts of key information packets per unit time is extracted from the database. If the number of broadcasts of critical information packets exceeds the broadcast frequency threshold within a unit of time, a downgrade is triggered. The specific process is as follows: some critical information packets are switched to a batch transmission mode via the broadcast link, and the non-critical aggregation window is dynamically expanded. If the number of times the critical information packet is broadcast within a unit of time does not exceed the broadcast number threshold, then transmission continues.

9. The dual-mode switching control system for communication links based on cloud-edge collaboration and broadcast subscription as described in claim 1, characterized in that, The specific analysis method for the targeted strategy of generating abnormal information packets is as follows: After receiving the abnormal information packet in the cloud, the global feature library is called to perform multi-dimensional matching; If the abnormal information package is successfully matched with the environmental drastic change feature sub-database, the cloud generates an abnormal information package with a targeted strategy of matching analysis threshold adjustment. The device corresponding to the abnormal information packet is identified and denoted as the target device. Determine the historical cycle length of the target device based on matching anomaly indicators; Select the target device's status pattern within a historical period. If the comparison with the target device's historical status pattern is successful; Obtain the similarity between the anomaly information packet and each historical state pattern to obtain the anomaly information packet similarity set; Extract the preset similarity threshold from the database; Traverse the set of similarity scores for abnormal information packets, select the maximum similarity score, and compare the maximum similarity score with the similarity threshold. If the maximum similarity is greater than or equal to the similarity threshold, then preventative maintenance is activated. If the maximum similarity is less than the similarity threshold, an early warning message is generated, and the system immediately downgrades to a lower level.

Images