Upgrading Method for Mine Intelligent Control System Supported by Downhole Ring Network

Abstract

The present invention discloses a method for upgrading a mine intelligent control system supported by an underground ring network, which relates to the technical field of data processing. The method includes: Step 1: Deploy edge computing nodes underground, and each edge computing node corresponds to an independent functional module of the mine intelligent control system; Step 2: When a certain functional module needs to be upgraded, the upgrade data is sent to the underground ring network, and the underground ring network controls the upgrade data through the SDN controller according to the header in the upgrade data, and transmits it to the edge computing node corresponding to the functional module via the data link; Step 3: After receiving the upgrade data, the edge computing node isolates the functional module from the mine intelligent control system in the upgrade part; Step 4: After the upgrade is completed, the upgrade part isolates the twin part from the mine intelligent control system. The present invention significantly improves the upgrade success rate and data transmission stability of the mine intelligent control system.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and particularly to a method for upgrading a mine intelligent control system supported by an underground ring network. Background Art

[0002] The mine intelligent control system is an important part of modern mining production. Its core goal is to achieve efficient, safe, and stable operation of mine production through automated, informatized, and intelligent technical means. With the deep mining of coal, metal mines, and non-metal mines, the mine environment has become increasingly complex, involving a large amount of monitoring data, a complex control system, and growing network communication requirements. Against this background, the need for upgrading the mine intelligent control system has become increasingly urgent. However, the existing technologies still face many challenges in the process of upgrading the mine control system, including instability of network communication, downtime risk during system upgrade, data loss, low path switching efficiency, function failure during upgrade, and energy consumption problems.

[0003] Currently, the mine intelligent control system mainly consists of functional modules such as mine environment monitoring, equipment automatic control, production scheduling and optimization. The core technologies include industrial Internet of Things (IIoT), edge computing, software-defined network (SDN), wireless sensor network (WSN), 5G communication, and cloud computing. Although the application of these technologies has significantly improved the level of mine automation and informatization, there are still obvious limitations in system upgrade. Currently, the networks in the mine environment mainly adopt wireless communication technologies such as fiber optic communication, industrial Ethernet, Wi-Fi 6, 5G private network, and LoRa. However, due to the complex terrain inside the mine, factors such as rock formation structure changes, high humidity, and high dust underground seriously interfere with signal transmission, resulting in an increase in data link delay, an increase in packet loss rate, and frequent link interruptions. During the upgrade process of the mine intelligent control system, if the instability of the transmission channel is not effectively solved, the upgrade data packets may be lost or the transmission may fail, resulting in the failure of system upgrade and affecting mine production safety. Summary of the Invention

[0004] The present disclosure is to provide a method for upgrading a mine intelligent control system supported by an underground ring network. The present invention adopts twin computing and module isolation mechanisms to ensure that system functions continue to operate during the upgrade, and at the same time combines shared self-repair vectors and consistent hash mapping to improve the stability, low latency and load balancing of path switching. In addition, the present invention uses data link performance prediction and least squares evolution trend analysis to perceive link degradation in advance and perform preventive path switching, effectively reducing the risk of data loss, and adds energy consumption optimization strategies during path selection, giving priority to low-power paths to improve the energy utilization efficiency of the mine network. The present invention significantly improves the upgrade success rate, data transmission stability, security and green energy-saving level of the mine intelligent control system, and provides key technical support for the intelligent, unmanned and long-term stable operation of mines.

[0005] The technical solution of the present invention is achieved in this way:

[0006] A method for upgrading a mine intelligent control system supported by an underground ring network, the method comprising:

[0007] Step 1: Deploy edge computing nodes underground. Each edge computing node corresponds to an independent functional module of the mine intelligent control system. Each edge computing node is connected to the underground ring network. Each edge computing node consists of a twin part and an upgrade part. The twin part is a real-time mirror of the functional module corresponding to the edge computing node.

[0008] Step 2: When a functional module needs to be upgraded, the upgrade data is sent to the underground ring network. Based on the header in the upgrade data, the underground ring network controls the upgrade data through the SDN controller, controls the routing, and transmits it to the edge computing node corresponding to the functional module via the data link;

[0009] Step 3: After receiving the upgrade data, the upgrade part of the edge computing node isolates the functional module from the mine intelligent control system, and the twin part replaces the functional module to implement the corresponding function in the mine intelligent control system; the upgrade part transmits the upgrade data to the isolated functional module for upgrade;

[0010] Step 4: When the upgrade is completed, the upgrade part isolates the twin part from the mine intelligent control system and releases the isolation of the functional module, allowing the functional module to be reconnected to the mine intelligent control system.

[0011] Furthermore, the functional modules of the mine intelligent control system include at least: a monitoring and sensing module for real-time monitoring of the mine environment, equipment operating status and personnel safety, a data acquisition and communication module for collecting, transmitting and storing underground data, an intelligent control and automation module for automatic control of various types of mine equipment, and a production optimization and intelligent scheduling module for scheduling and optimizing mine production processes.

[0012] Furthermore, when a device failure or power supply interruption occurs in the downhole ring network due to the downhole environment, causing a failure in a certain data link or route, the SDN controller executes a self-healing routing process to ensure that the upgrade data is sent to the corresponding functional module under the control of the route through other data links; the route includes: a primary route and a backup route.

[0013] Furthermore, the SDN controller's execution of the self-healing routing process specifically includes: the SDN controller maintains the global topology information and data link status data structure for multiple edge computing nodes. The data link status data structure includes an edge computing node reachability identification matrix and an edge computing node inter-data link performance vector. The edge computing node reachability identification matrix orthogonally identifies the current active state of each data link or edge computing node for the paths between multiple edge computing nodes. The data link performance vector includes one or more of delay, bandwidth, and packet loss rate; the SDN controller periodically calculates and pre-stores a self-healing backup path list based on the data link status data structure. The backup path list includes several backup routes for each primary route path, and each path entry in the backup path list includes a path sequence identifier, a path availability probability vector, and a path switching cost vector; the SDN controller distributes the pre-computed backup path list to the virtual router proxy edge computing nodes it manages.

[0014] Furthermore, the virtual router proxy edge computing node maintains a local path switching forwarding table based on the backup path list. The local path switching forwarding table includes a shared self-healing vector orthogonal to the primary path to dynamically mark the real-time state of each backup path. The shared self-healing vector is shared by multiple virtual router proxy edge computing nodes and is used to identify the active or to-be-activated state of each path in the backup path set.

[0015] Furthermore, in response to the SDN controller detecting a failure in one or more data links in the primary path, the SDN controller notifies the corresponding virtual router proxy edge computing node to update the shared self-healing vector in real time; in response to the shared self-healing vector identifying the backup path state as the active state, the virtual router proxy edge computing node dynamically executes a network service switching process to seamlessly switch the network service from the primary path where the failed data link is located to the optimal backup path in the backup path list and updates the local path switching forwarding table to reflect the current path state used by the traffic flow.

[0016] Further, the SDN controller's execution of the self-healing routing process further includes: the virtual router proxy edge computing node periodically sends a data link status detection message to the SDN controller, and the data link status detection message carries a data field identifying the real-time performance of the data link on the current path; in response to the SDN controller receiving the data link status detection message, the SDN controller updates the data link performance vector and, based on the updated data link performance vector, recalculates the path switching cost vector in the standby path list in real time; when the updated path switching cost vector exceeds a preset threshold, the SDN controller actively triggers the virtual router proxy edge computing node to perform a path pre-switching action and pre-switch to a standby path with a lower cost in the standby path list.

[0017] Further, the dynamic execution of the network service switching process includes: the virtual router proxy edge computing node maintains a consistent hashing mapping for the currently active network service path, and the consistent hashing mapping maps the network service flow identifier to a set of edge computing nodes on the hashing ring; in response to changes in the activation or pending activation state of the path status, the consistent hashing ring is dynamically updated to exclude the edge computing nodes or data links corresponding to the faulty data link, and the transmission path of the network service flow is remapped to ensure the uniform distribution and load balancing of the network service flow; during the service switching process, the continuity of the original service flow and the hashing ring mapping is maintained to avoid service jitter caused by frequent path switching in a short period of time.

[0018] Further, the delay and packet loss rate in the data link performance vector are processed by the SDN controller using a statistical sliding window algorithm, and the least squares method is used to predict the evolution trend of the data link performance, which is used to predict the probability of data link failure; when the predicted evolution trend of the data link performance exceeds the set warning threshold, the SDN controller actively issues a preventive standby path switching instruction to the virtual router proxy edge computing node in advance.

[0019] Further, the self-healing vectors are shared in real time among multiple virtual router proxy edge computing nodes through a multicast channel, and a message queue and a distributed consistency protocol are used to ensure the consistency of the path switching state; the path switching cost vector further includes the energy consumption index of the path, and the SDN controller preferentially selects a standby path with lower energy consumption for switching based on the path energy consumption prediction to achieve the goal of network green energy conservation.

[0020] The method for upgrading the mine intelligent control system supported by the underground ring network of the present invention has the following beneficial effects: The mine environment is complex, and the data link is easily affected by factors such as rock formation structure, humidity, dust, and electromagnetic interference, resulting in data transmission interruption or packet loss. In the traditional mine system upgrade method, once the data link is unstable, the upgrade data transmission may fail, causing the upgrade process to be forced to interrupt, and even the system needs to be restarted. The present invention adopts the cooperation of the underground ring network and the SDN controller. Through the data link status detection message and the statistical sliding window algorithm, it real-time monitors the health status of the data link, and uses the least squares method to predict the evolution trend of the data link. When it is found that the performance degradation trend of a certain data link exceeds the set threshold, the SDN controller will adjust the transmission path of the data stream in advance to ensure that the upgrade data can be continuously and stably transmitted, avoiding the failure of the upgrade process due to link failures. Most of the traditional mine intelligent control system upgrades rely on offline upgrades or shutdown maintenance, resulting in a decrease in mine production efficiency and even an increase in production safety risks. The present invention adopts the twin computing technology, that is, the twin part and the upgrade part are deployed on the edge computing node to ensure that during the upgrade of the functional module, the twin part can take over the original function and ensure the continuous operation of the mine control system. For example, during the upgrade of the ventilation system, the twin part can simulate the fan control logic and continuously adjust the operating state of the fan during the upgrade to ensure that the mine ventilation system will not fail due to the upgrade, thus avoiding safety hazards such as mine gas accumulation. The online hot upgrade method of the present invention enables the mine intelligent control system to complete the upgrade without interrupting production, improving system availability. In the traditional mine network, path switching usually relies on static routing or simple backup path mechanisms, resulting in slow switching speed and possible short-term loss of data streams when link failures occur, affecting the continuity of the upgrade. The present invention combines the SDN controller with the virtual router to proxy the edge computing node, and adopts the shared self-healing vector and multicast channel to ensure that all edge computing nodes can synchronously perceive the change of the path state. In addition, the present invention introduces a consistent hashing mapping mechanism to ensure that during path switching, the traffic flow can be evenly distributed to the optimal backup path and maintain the continuity of the original traffic mapping during the switching process, thus avoiding service jitter caused by frequent path switching. This mechanism significantly improves the efficiency of path switching, enabling the system to complete path adjustment within milliseconds to ensure the stability and efficiency of upgrade data transmission. The present invention adopts a dynamic path optimization strategy, enabling the SDN controller to dynamically adjust the transmission path of the data stream according to the delay, bandwidth, packet loss rate, and load conditions of the path during the upgrade process. Different from the traditional static routing method, the present invention uses the path switching cost vector, combines the historical network state data and the current real-time load conditions, and intelligently selects the optimal path for data transmission.For example, among multiple available paths, if the load of a certain link is high, the SDN controller will preferentially select a low-load path to ensure that the upgrade data can be transmitted with the lowest latency and high bandwidth, thereby improving the upgrade success rate. In addition, the load balancing strategy of the present invention can also prevent the upgrade data transmission from being slow due to overload on certain links, improving the overall performance of the system. Description of the Drawings

[0021] Figure 1 It is a schematic flowchart of the method for upgrading a mine intelligent control system supported by an underground ring network provided by an embodiment of the present invention. Detailed Embodiments

[0022] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present disclosure clearer and more understandable, the present disclosure will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.

[0023] Embodiment 1, refer to Figure 1 : A method for upgrading a mine intelligent control system supported by an underground ring network, the method comprising:

[0024] Step 1: Deploy edge computing nodes underground, each edge computing node corresponding to an independent functional module of the mine intelligent control system; each edge computing node is connected to the underground ring network; each edge computing node includes: a twin part and an upgrade part; the twin part is a real-time mirror of the functional module corresponding to the edge computing node.

[0025] The mine intelligent control system involves many subsystems, such as mine pressure monitoring, ventilation control, gas concentration detection, shearer control, and personnel safety management. The operating environments of these subsystems are extremely complex, and it is necessary to ensure the real-time and stable data processing. Therefore, deploying edge computing nodes underground, so that each edge computing node corresponds to an independent functional module, thereby forming a decentralized computing architecture, can effectively solve the problems of data transmission delay, bandwidth bottleneck, and single-point failure risk caused by centralized computing in traditional mine control systems. In traditional mine control systems, all data is usually collected by various sensors and then transmitted to a central control server for processing. Due to the closed nature of the mine environment and the limitations of network conditions, the reliability of data transmission is difficult to guarantee. Moreover, when the mine scale is large, the data transmission path is long and the computing pressure is concentrated, which easily leads to system response lag and even inability to respond in a timely manner in case of emergencies. The present invention adopts an edge computing architecture, supported by the underground ring network, to sink the data processing ability to the underground site. Each edge computing node independently undertakes the tasks of data collection, processing, and decision-making for a functional module, and realizes data interaction between each computing node through the underground ring network, making the entire system have higher real-time and reliability.

[0026] The edge computing node consists of two parts, namely, the twin part and the upgrade part. The twin part is used to maintain the real-time mirror of the function module corresponding to this edge computing node, while the upgrade part is responsible for the isolation and recovery of the function module during the system upgrade process. The introduction of the twin part enables the operating state of the original function module to be completely replicated to the twin part during the upgrade process, thus maintaining the normal operation of the system during the upgrade without affecting the continuity of mine production. Based on this characteristic of the twin part, the operation of the mine intelligent control system no longer depends on a single computing node. Even if a certain function module is temporarily offline due to upgrade or maintenance, the twin part can still maintain the execution of this function, thus ensuring the stability of the system. Each edge computing node is connected to the underground ring network to form an extensible distributed computing network. In the traditional mine network architecture, the data flow often adopts a fixed routing mode, and the network topology is relatively static, which is not suitable for the dynamic changes of the mine environment. However, the present invention utilizes the flexible architecture of the underground ring network to enable the edge computing nodes to be dynamically connected and perform traffic scheduling and path optimization through an SDN (Software Defined Network) controller to ensure the data transmission efficiency. By integrating with the underground ring network, the edge computing node not only enhances the computing power of the mine control system but also enables the upgrade data to be efficiently transmitted in the network, supporting the realization of the online upgrade function.

[0027] The introduction of edge computing nodes also effectively reduces the bandwidth occupancy of the system. Since the computing tasks are completed locally, only the key decision-making data needs to be uploaded to the central control system, instead of transmitting all the original data to the ground for processing. This method greatly reduces the network load and improves the data processing efficiency at the same time. For example, in the gas concentration detection system, in the past, all sensor data needed to be transmitted to the ground server for calculation. After adopting the method of the present invention, the edge computing nodes can complete data preprocessing locally, such as filtering, anomaly detection and trend analysis, and only send alarm signals to the central system when anomalies are detected. This method not only improves the system response speed, but also avoids unnecessary data transmission. In addition, the deployment of edge computing nodes also provides guarantee for the security of the system. Due to the uncertainties in data transmission in the mine environment, such as electromagnetic interference, signal attenuation and other problems, the centralized computing mode may lead to data loss or computing interruption due to network fluctuations. After adopting the distributed computing architecture, even if some edge computing nodes fail, other nodes can still maintain the normal operation of the system, avoiding the paralysis of the entire mine control system caused by single-point failure. At the same time, the design of the twin part not only ensures the stable operation of the functional module during the upgrade process, but also provides guarantee for the abnormal recovery of the system. When a functional module needs to be repaired or rolled back due to a fault, the twin part can temporarily replace the original functional module to ensure that the system can still operate normally during the repair process. Through the support of the underground ring network, the edge computing nodes of the present invention can be flexibly accessed to form a decentralized and dynamically scalable computing architecture, thus solving the problems of communication delay, data transmission bottleneck and insufficient system availability brought by the centralized computing mode of the traditional mine intelligent control system. At the same time, the separated design of the twin part and the upgrade part provides support for online upgrade, enabling the mine intelligent control system to complete the upgrade and maintenance of functional modules without interrupting production.

[0028] Step 2: When a certain functional module needs to be upgraded, send the upgrade data to the underground ring network. The underground ring network controls the upgrade data according to the header in the upgrade data, and through routing control, transmits the upgrade data to the edge computing node corresponding to the functional module via the data link;

[0029] Due to the particularity of the mine environment, the stability and low-latency characteristics of network communication are crucial. Most traditional mine networks adopt static routing methods, that is, the data transmission path is preset in advance during network configuration, resulting in the inability to dynamically adjust the transmission path in the face of emergencies or network state changes, which may further lead to data congestion, increased latency, and even network interruption. To overcome these limitations, the present invention adopts an underground ring network architecture and combines it with an SDN controller for intelligent scheduling of data streams, enabling the upgrade data to be quickly and accurately delivered to the target edge computing node, providing technical guarantee for the online upgrade of the system.

[0030] When a certain functional module needs to be upgraded, the upgrade data will first be sent to the underground ring network. The underground ring network does not directly perform broadcast transmission, but rather performs precise routing based on the policies of the SDN controller. The key to this mechanism lies in the packet header of the upgrade data, which contains the target address, data type, priority information, etc. of the upgrade data. The SDN controller will parse the header information and dynamically plan the data transmission path according to parameters such as the current topology of the underground network, bandwidth status, and link load conditions, ensuring that the upgrade data can reach the target edge computing node with the lowest latency and optimal reliability. Compared with the traditional static routing method, the introduction of the SDN controller makes the mine network highly flexible. Even if a certain data path becomes congested or some network nodes fail, the controller can still adjust the routing policy in real time and select the optimal transmission path, thus avoiding upgrade failures caused by network anomalies. During the transmission process of the upgrade data, the optimization of the data link layer is also the key to ensuring efficient transmission. Due to the complex mine environment, factors such as signal attenuation, interference, and the physical distance between nodes will all affect the data transmission quality. The data transmission of the traditional mine network often relies on communication channels with fixed bandwidths. However, the underground ring network of the present invention combines the bandwidth management capabilities of the SDN controller and can achieve intelligent scheduling at the data link layer. Specifically, the SDN controller can dynamically allocate bandwidth resources according to the urgency of the upgrade data and the network load conditions, ensuring that the upgrade data has a higher transmission priority and optimizing the transmission protocol to reduce the possibility of packet loss. In this way, even when the mine network load is high, the upgrade data can still be preferentially processed, ensuring the smooth progress of the upgrade task.

[0031] When the upgrade data reaches the target edge computing node, the integrity and security of the data are another important consideration. The present invention adopts a data transmission verification method based on a verification mechanism to ensure that the data will not be damaged due to channel interference or packet loss during the transmission process. During the data transmission process of the underground ring network, each data packet is attached with a hash verification code. After receiving the data packet, the receiving edge computing node will perform a hash verification on the data and compare it with the verification code of the sending end to determine whether the data has been tampered with or damaged during the transmission process. If the verification fails, the edge computing node will request the retransmission of the corresponding data packet to ensure the integrity of the upgrade data. In addition, to prevent the data from being maliciously tampered with or attacked, the underground ring network of the present invention adopts an encrypted transmission mechanism to ensure that the upgrade data will not be illegally tampered with or intercepted due to external interference. The mine intelligent control system is related to the safety and production operation of the mine. Any malicious data tampering may pose a serious threat to the mine safety. Therefore, end-to-end encryption is adopted during the data transmission process, further enhancing the security of the system. The underground ring network and the SDN controller of the present invention work together to make the transmission process of the upgrade data more efficient, intelligent and secure. The traditional upgrade method of the mine intelligent control system usually relies on manual operation. During the upgrade, the network path needs to be manually configured one by one, and in some cases, the machine needs to be shut down for data update. However, the present invention realizes the accurate transmission of the upgrade data through automated data scheduling and intelligent traffic management, thus greatly improving the upgrade efficiency of the system. In addition, due to the underground ring network having a ring topology structure, in the case of partial node failures, the data can still be transmitted through the backup path, and this redundancy mechanism further improves the availability of the mine intelligent control system.

[0032] In practical application scenarios, such as the upgrade of the shearer control system, the SDN controller can judge the priority of data transmission according to the current operating state of the shearer. If the shearer is running at high speed, the SDN controller can postpone the upgrade of non-critical function modules and give priority to the bandwidth resources for the real-time transmission of production data, thus ensuring the continuity of mine production. When the shearer is in a low-load state, the SDN controller can dynamically adjust the bandwidth allocation strategy and give priority to transmitting the upgrade data to ensure that the upgrade can be completed as soon as possible. This flexible network scheduling method avoids the network congestion problems that may be caused by traditional upgrade methods and also ensures that the upgrade task can be completed along the optimal path.

[0033] Step 3: After receiving the upgrade data, the upgrade part of the edge computing node isolates the function module from the mine intelligent control system, and the twin part replaces the function module to realize the corresponding function in the mine intelligent control system; the upgrade part transmits the upgrade data to the isolated function module for upgrade;

[0034] The intelligent control system of the mine involves multiple key functional modules, which are responsible for controlling and monitoring the environment, equipment, and personnel safety in the mine. During the upgrade process, if the functions are directly replaced, it may lead to a short-term interruption of the system and even the risk of control failure. Therefore, the present invention adopts an upgrade strategy based on twin computing, that is, the twin part of the edge computing node temporarily takes over the work of the functional module, so as to ensure that the system can still operate normally during the upgrade process, while the upgraded part is responsible for isolating the original functional module and executing the upgrade task, thus realizing seamless switching.

[0035] In the upgrade mechanism of the present invention, the twin part of the edge computing node is equivalent to a dynamic mirror of the functional module, which can synchronize the running state of the functional module in real time and take over the work of the module when necessary. At the beginning of the upgrade, the edge computing node first isolates the functional module to be upgraded, that is, strips the real-time control authority of the module from the intelligent control system of the mine to ensure that other system functions will not be affected during the upgrade process. The key to the isolation process lies in ensuring the integrity and consistency of the data. When isolation occurs, the edge computing node needs to ensure that the twin part has fully synchronized the running state of the current functional module and can independently execute all the computing tasks of the module, so that the whole system can not perceive the switching of the functional module, thus avoiding affecting the continuity of mine production due to the short-term instability during the switching process. For example, in the mine ventilation control system, if the fan control module needs to be upgraded and the module is directly deactivated for upgrade, it may lead to a short-term failure of the ventilation system and pose a safety hazard, while the solution of the present invention can have the twin part take over the fan control during the upgrade process to ensure the stable operation of the whole ventilation system.

[0036] After isolation is completed, the upgrade part is responsible for transmitting the upgrade data to the isolated functional module and executing the upgrade task. Since functional modules usually involve complex calculation logics and real-time control parameters, the upgrade process needs to ensure the integrity of data and the correctness of algorithms. Therefore, the present invention designs an intelligent upgrade mechanism in the upgrade part, that is, integrity verification of the upgrade data is performed before the upgrade, and the system status is monitored in real time during the upgrade to ensure that the upgraded functional module can correctly resume operation. The traditional upgrade of mine intelligent control systems usually adopts an offline method, that is, the system is upgraded during mine shutdown and maintenance, while the online upgrade method of the present invention enables the mine control system to update software and functional modules without interrupting production through twin calculation and intelligent isolation mechanisms, greatly improving the availability and safety of the system. During the upgrade process, the twin part not only simply substitutes for the functional module to perform calculation tasks, but it can also collect the operation data of the mine intelligent control system in real time and make dynamic adjustments to adapt to the real-time changes of the system. This means that during the upgrade of the functional module, the twin part can dynamically adjust the control strategy based on the existing control model and real-time data to ensure that the system does not malfunction due to external environmental changes during the upgrade. For example, when the mine pressure monitoring system is upgraded, if the stress state inside the mine changes suddenly, the twin part can make dynamic adjustments based on the existing sensing data to ensure that the mine pressure monitoring data remains reliable during the upgrade, thus avoiding safety risks caused by system upgrade.

[0037] After the upgrade part completes the update of the functional module, the system needs to perform status verification to ensure that the upgraded functional module can operate correctly and can be seamlessly connected to the mine intelligent control system. In the traditional mine control system upgrade method, the upgraded module usually needs manual testing and debugging before it can be put into use again. The present invention adopts an automatic regression verification mechanism, that is, after the upgrade is completed, the upgrade part will verify the behavior of the new functional module through the historical data of the twin part to ensure that its calculation logic and control strategy are consistent with the operating state before the upgrade, so as to avoid sudden changes in system parameters or abnormal control logic due to the upgrade. If the verification is passed, the upgrade part will release the isolation of the functional module, so that it can be reconnected to the mine intelligent control system, and take over the tasks of the twin part to complete the formal functional handover. At this time, the twin part gradually exits the working mode and returns to the standby state so that it can be enabled again at the next upgrade. During the entire upgrade process, the real-time performance and reliability of the mine intelligent control system are always guaranteed. This is an important breakthrough of the present invention compared to the traditional mine control system upgrade method. Traditional upgrade methods usually require downtime for maintenance, or operate in a degraded mode during the upgrade, which limits some functions of the system. The present invention uses twin computing and intelligent isolation technology to enable the system to complete the upgrade under full-load operation, ensuring the continuity and safety of mine production. In addition, the upgrade mechanism of the present invention is highly scalable. Even in a complex mine environment, multiple functional modules can be upgraded at the same time without affecting each other, because each edge computing node can independently perform the upgrade task and achieve efficient data transmission and status synchronization through the underground ring network. The online upgrade method of the present invention is applicable to a variety of mine control system scenarios. For example, when the remote control system of underground electromechanical equipment is upgraded, the twin part can continuously monitor the equipment status during the upgrade and perform early warning control to prevent equipment failures due to short-term abnormalities during the upgrade process. When the personnel safety management system is upgraded, the twin part can continuously track the miners' location information during the upgrade process to ensure that the personnel safety monitoring function will not be interrupted due to the upgrade. This design greatly improves the flexibility of the mine intelligent control system, enabling the system to perform intelligent upgrades in highly complex and dynamically changing environments.

[0038] Step 4: When the upgrade is completed, the upgrade part isolates the twin part from the mine intelligent control system and releases the isolation of the functional module, allowing the functional module to be reconnected to the mine intelligent control system.

[0039] In the previous step, the functional module has completed the upgrade in an isolated state, and the continuous and stable operation of the system during the upgrade period is ensured by the takeover of the twin part. The key task of step 4 is to remove the replacement effect of the twin part, so that the upgraded functional module can be reconnected to the mine intelligent control system, while ensuring that the control logic, data flow and functional coordination of the entire system will not be affected by the changes during the upgrade process. This process is not just a simple function switch, but a complex process involving system integrity verification, state consistency recovery, real-time data synchronization and dynamic switching control to ensure that the mine intelligent control system can quickly restore normal operation after the upgrade is completed and ensure its long-term stability. In the actual system operation process, the reconnection of the functional module after the upgrade is a highly sensitive link, because the operating state of the upgraded module may be slightly different from the original state, and if these differences are not strictly controlled, they may cause system instability or even operation errors. Therefore, before the twin part exits, the present invention will first perform system state comparison and consistency verification to ensure that the upgraded functional module is consistent with the operating results of the twin part in terms of data input, logic processing, control instruction execution, etc. In the process of the twin part temporarily replacing the functional module to perform tasks, it has accumulated a certain amount of system data, which will serve as a benchmark for state comparison. The upgraded functional module needs to go through this comparison process before taking over to ensure that all key variables, system parameters and control states are correctly connected, thereby avoiding system abnormalities due to state drift or data mutation. For example, during the upgrade of the mine pressure monitoring system, the twin part continuously collects underground mine pressure data during the upgrade period and speculates on future trends through prediction algorithms. Before the upgraded functional module is reconnected to the system, it needs to ensure that its calculation results are consistent with the twin part's prediction results. Otherwise, it may cause monitoring data mismatch, thereby affecting the accuracy of mine safety management.

[0040] After the completion of the status comparison, the upgrade part issues a switching instruction to gradually transfer the control right of the twin part to the upgraded functional module. This process is not completed instantaneously, but a progressive dynamic adjustment process, which is carried out under the condition of relatively stable system operation state to avoid the impact of sudden switching on the entire mine intelligent control system. The present invention adopts a progressive dynamic switching method, that is, the twin part and the upgraded functional module are simultaneously run within a short period of time, so that the control signals, data streams and calculation logics of the system can smoothly transition between the two, rather than a sudden direct switching. For example, during the upgrade of the underground ventilation control system, the twin part controls the operation parameters of the fan during the upgrade period, and when the upgraded functional module is reconnected, it does not directly take over all the control tasks of the fan, but first runs in a low-priority mode and gradually increases its control weight until it completely replaces the control logic of the twin part, so as to ensure that the entire fan system will not have abnormal fluctuations due to sudden switching. In addition to the function switching, data synchronization is also a key link in the reconnection of the functional module. During the upgrade period, there may be differences between the twin part and the upgraded functional module in terms of data storage, log record and status caching. Therefore, before the twin part exits, the system needs to perform data synchronization to ensure that all relevant historical data can be correctly transmitted to the upgraded functional module and stored and updated inside it. The upgrade methods of traditional mine control systems usually do not pay attention to this detail, so data mismatch may occur after the upgrade, resulting in the system needing to re-learn and adapt to new parameters in a short period of time. However, the design of the present invention enables the upgraded functional module to immediately obtain complete historical operation data through the data synchronization mechanism, so that there is no need for an additional adaptation period when taking over the task again, improving the recovery speed and stability of the system.

[0041] After completing data synchronization and function switching, the twin part will gradually exit the system and return to the standby mode for reuse during future upgrades. The design of the twin part enables the entire system to be upgraded online at any time without affecting the normal operation of the mine intelligent control system, while also ensuring the long-term stability of the system. After the upgrade is completed, the system will also perform a final integrity check to ensure that the upgraded function modules can execute all instructions correctly and meet the requirements of mine production and safety management. If any abnormalities are detected at this stage, the system can immediately activate the rollback mechanism, allowing the twin part to resume tasks and troubleshoot the upgraded function modules without affecting the normal operation of the mine. This intelligent rollback mechanism further enhances the system's security and avoids potential production losses caused by upgrade failures. The upgrade method of the present invention breaks through the traditional upgrade mode of mine intelligent control systems, enabling the mine to perform smooth system upgrades under high-load operating conditions without affecting mine safety production. Through the dynamic takeover of the twin part, the progressive switching of the upgraded function modules, the data synchronization mechanism, and the intelligent rollback mechanism, the present invention constructs a highly stable and efficient online upgrade system to ensure that the mine intelligent control system always maintains optimal performance during long-term operation and can quickly adapt to the iteration and update of new technologies and new functions. In the complex underground environment, this upgrade method not only improves the system's reliability but also reduces the potential safety risks during the upgrade process, enabling the intelligent management of the mine to operate more stably and efficiently.

[0042] Embodiment 2: The function modules of the mine intelligent control system at least include: a monitoring and sensing module for real-time monitoring of the mine environment, equipment operating status, and personnel safety, a data acquisition and communication module for underground data acquisition, transmission, and storage, an intelligent control and automation module for automatic control of various mine equipment, and a production optimization and intelligent scheduling module for scheduling and optimizing the mine production process.

[0043] Specifically, first of all, the monitoring and perception module is the basic layer of the mine intelligent control system, responsible for real-time monitoring of the mine environment, equipment operation status, and personnel safety. The internal environment of the mine is complex, involving multiple key safety factors such as gas concentration, dust concentration, temperature and humidity, wind speed and wind pressure, mine pressure, and water hazard monitoring. Therefore, the main function of the monitoring and perception module is to collect these environmental parameters in real time through high-precision sensors and give early warnings for abnormal situations. At the same time, this module is also responsible for monitoring the operation status of equipment, including the working status, energy consumption level, and fault prediction of key equipment such as shearers, conveyor belts, fans, and pumps. In addition, the safety monitoring of miners is also an important part of this module. For example, through technologies such as RFID, UWB ultra-wideband positioning, and video monitoring, the position information of miners can be tracked in real time to ensure that mine personnel can be rescued quickly in case of emergency. The traditional mine monitoring system relies on a single sensor or manual inspection. The monitoring and perception module of the present invention adopts multi-sensor fusion technology to achieve high-precision data collection and can maintain the continuity of the monitoring function through twin computing technology during the system upgrade process to ensure that the monitoring data will not be interrupted during the upgrade period.

[0044] Secondly, the data acquisition and communication module is the data center of the mine intelligent control system, responsible for collecting, transmitting, and storing various data information underground. The present invention combines edge computing nodes with the underground ring network to achieve distributed data collection and transmission, avoiding the network delay and bandwidth bottleneck problems caused by centralized data processing in the traditional mine monitoring system. In terms of data transmission, the data acquisition and communication module supports a combination of wired communication (such as industrial Ethernet, optical fiber) and wireless communication (such as 5G, LoRa, self-organizing network), enabling data to be reliably transmitted under different network conditions. Especially during the upgrade process, the stability of the data acquisition and communication module is directly related to the continuity of the mine intelligent control system. Therefore, in the upgrade method of the present invention, during the upgrade process of this module, it will first ensure the stable operation of the data link to avoid data loss or communication interruption caused by system upgrade. In addition, this module also has the functions of data storage and distribution, and can locally store key data at the edge computing node, reducing the dependence on the central server and improving the anti-interference ability of the system.

[0045] The intelligent control and automation module is the execution layer of the mine intelligent control system, responsible for automating the control of various mine equipment to improve production efficiency and safety. The core tasks of this module include remote control, automatic adjustment, and intelligent optimization of key equipment such as shearers, hydraulic supports, belt conveyors, ventilation systems, and water pumps. In the traditional mine production mode, the control of equipment mainly relies on manual operation or simple PLC (programmable logic controller) logic. This invention realizes adaptive control through intelligent control algorithms combined with sensor data. For example, in the ventilation control system, this module can automatically adjust the fan speed according to the real-time gas concentration and miner distribution to ensure safe ventilation and reduce energy consumption. During the system upgrade process, the upgrade of the intelligent control and automation module may involve algorithm optimization and the deployment of new control logic. Therefore, this invention adopts twin computing technology, enabling the upgrade process to be simulated and tested in a virtual environment and then formally applied after verification to ensure that the upgrade does not affect the stable operation of the equipment.

[0046] Finally, the production optimization and intelligent scheduling module is the decision-making layer of the mine intelligent control system, responsible for optimizing the mine production process and intelligent scheduling. Based on mine environment monitoring data, equipment operation data, and personnel location information, this module uses artificial intelligence, big data analysis, and optimization algorithms to achieve automatic allocation of production tasks, resource optimization, and job scheduling. For example, this module can automatically schedule the operation rhythm of shearers and transportation systems according to the mine exploitation plan and equipment working status, avoiding efficiency losses caused by equipment waiting or unreasonable resource scheduling during the production process. In addition, this module can dynamically adjust the operation plan. For example, after detecting that the gas concentration in a certain area underground exceeds the standard, it automatically adjusts the operation area to ensure the safety of mine production. Traditional mine production scheduling relies on manual decision-making, often lacking real-time and intelligence. This invention enables the mine production to achieve the highest efficiency under the premise of ensuring safety through the intelligent optimization algorithm of this module. During the upgrade process, the computing tasks of the production optimization and intelligent scheduling module are relatively complex. Therefore, the upgrade method needs to minimize the impact of the upgrade process on the mine production plan. To this end, during the upgrade process of this module, this invention ensures that the scheduling logic can remain consistent during the upgrade through the dynamic mapping of the twin part, and performs status verification after the upgrade to ensure the continuity and correctness of the scheduling strategy.

[0047] Embodiment 3: When the underground ring network fails or the power supply is interrupted due to the underground environment, resulting in a failure of a certain data link or route, the SDN controller executes a self-healing routing process to ensure that the upgrade data is sent to the corresponding functional module under the control of the route through other data links; the route includes: a main route and a backup route.

[0048] Compared with the ground environment, the mine environment has many unstable factors, such as humidity, high temperature, high dust, geological movement, etc. These factors may cause damage to underground communication equipment, power supply interruption, and even failure of some network nodes, thus affecting the normal operation of the data link. Traditional mine networks mostly adopt static routing strategies, and the transmission paths of data streams are preset during network deployment. Once a data link fails, the communication ability of the entire network will be greatly reduced, and even the entire intelligent control system will be paralyzed. However, the SDN controller of the present invention can still ensure the stable transmission of data streams in the case of partial failure of the mine network by real-time monitoring of the network status and performing self-repair routing processes, enabling the upgraded data to reach the target function module smoothly and ensuring that the upgrade process of the mine intelligent control system will not be interrupted due to network failures.

[0049] During the data transmission process, the SDN controller of the present invention adopts a combination of a primary route and a backup route to improve the network fault tolerance of the system. The primary route is the main transmission path for upgraded data. Under normal circumstances, the SDN controller will preferentially select the path with the best network status, sufficient bandwidth, and the smallest delay as the primary route to ensure that the data can be transmitted to the target function module with the highest efficiency. However, the complexity of the mine environment makes it difficult to guarantee the stability of the primary route for a long time. For example, when a switch fails due to a power supply fault, or a fiber optic cable breaks due to geological changes, the original primary route will no longer be able to undertake the data transmission task. Under the traditional mine network architecture, such faults usually require manual intervention for network reconfiguration. However, the SDN controller of the present invention adopts an intelligent self-repair mechanism that can immediately switch to the backup route the moment the primary route fails to ensure the continuity of the data link.

[0050] In the design of the present invention, the SDN controller configures multiple backup routes for each data stream in advance by continuously monitoring the link status of the underground network. The backup route is a path that can immediately take over the data stream transmission when the main route fails. Its selection criteria are not only based on the network topology, but also combined with real-time network status data, such as the current load of each link, bandwidth occupancy, network congestion level, etc. In a mine environment, the status of the data link changes dynamically. Therefore, when the SDN controller performs self-repair routing, it does not simply select a preset backup path, but adopts an intelligent path selection algorithm to dynamically calculate the optimal transmission solution among all available paths. For example, when the main route fails, the SDN controller will quickly evaluate all available backup paths and select the path with the lowest current latency and lightest load as the new main route to ensure that the upgraded data can be delivered to the target functional module quickly and stably. In actual application scenarios, such as the upgrade data transmission of a mine ventilation control system, if the SDN controller detects that the original main route is interrupted due to the failure of a network node, it will immediately start the self-repair mechanism and switch the upgrade data to another available backup route, so that the ventilation system can successfully complete the software upgrade without affecting the normal operation of the system due to network failure. This mechanism not only improves the availability of the mine intelligent control system, but also greatly reduces the risk of upgrade failure due to network failure. In some extreme cases, such as the collapse of a local area underground, resulting in large-scale network failure, the SDN controller of the present invention can also re-plan the data transmission path through the network reconstruction strategy, so that the upgrade data can bypass the damaged area and continue to complete the data transmission task, thereby ensuring the stability of the system upgrade.

[0051] In addition, during the execution of the self-healing routing process by the SDN controller of the present invention, a hierarchical recovery mechanism is adopted to minimize the impact of network failures on the mine intelligent control system. When a transient failure occurs in the primary route, for example, when a wireless communication link is temporarily interrupted due to mine electromagnetic interference, the SDN controller will first attempt to automatically restore the availability of this path through link retransmission technology. If it cannot be restored within a short period of time, it will immediately switch to the backup route. For a long-term failure of the primary route caused by hardware damage, the SDN controller will adopt a more advanced network topology reconstruction technology. By recalculating the network topology, it dynamically adjusts the transmission path of the data stream to adapt to the changes in the mine environment. This hierarchical recovery mechanism enables the mine intelligent control system of the present invention to have extremely high network adaptability. Even in a complex underground environment, it can maintain the stable operation of the system. Due to the lack of a dynamic routing mechanism, the network architecture of traditional mine intelligent control systems usually has difficulty coping with sudden network failures. However, the SDN controller of the present invention combines the primary route and the backup route and adopts an intelligent self-healing mechanism to achieve rapid recovery in the event of network failures, ensuring that the upgraded data of the mine intelligent control system can always remain available. Especially in an environment such as a mine where safety requirements are extremely high, the application of this mechanism enables system upgrades not to affect the normal operation of the mine due to network failures, ensuring the high reliability and high availability of the mine intelligent control system. Through the self-healing routing technology of the SDN controller, the present invention effectively breaks through the limitations of the traditional mine network architecture, provides a solid technical guarantee for the online upgrade of the mine intelligent control system, and also enables the mine intelligent management system to be more stable, efficient, and intelligent during long-term operation.

[0052] Embodiment 4: The specific process of the SDN controller executing the self-healing routing includes: The SDN controller maintains the global topology information and the data link status data structure for multiple edge computing nodes. The data link status data structure includes the edge computing node reachability identification matrix and the data link performance vector between edge computing nodes. The edge computing node reachability identification matrix orthogonally identifies the current active state of each data link or edge computing node for the paths between multiple edge computing nodes. The data link performance vector includes one or more of delay, bandwidth, and packet loss rate; The SDN controller periodically calculates and pre-stores the self-healing backup path list according to the data link status data structure. The backup path list includes several backup routes for each primary route path, and each path entry in the backup path list includes a path sequence identifier, a path availability probability vector, and a path switching cost vector; The SDN controller distributes the pre-calculated backup path list to the virtual router proxy edge computing nodes managed by it.

[0053] Specifically, the online upgrade of the mine intelligent control system relies on a reliable communication network. However, due to the influence of the mine environment, the underground ring network may experience the failure of some data links caused by equipment failures, electromagnetic interference, network congestion, power supply problems, or geological changes. Therefore, the SDN controller not only needs to have the ability to detect faults but also be able to dynamically calculate the optimal backup path to quickly switch the routing when the network is abnormal, ensuring that the upgrade data can reach the target function module with the lowest latency and the highest reliability. The self-healing routing process of the SDN controller first depends on its global perception ability of the entire underground network topology. For this purpose, the SDN controller maintains a global topology information and data link status data structure for multiple edge computing nodes. The core of the data link status data structure lies in the edge computing node reachability identification matrix and the data link performance vector. The reachability identification matrix is used to identify all edge computing nodes in the current underground ring network and the data link status between them. The construction method of this matrix is orthogonal, that is, it can uniquely determine the communication status between any two nodes in the mine network and can reflect the active status of a certain data link or an edge computing node in real time. When a data link is interrupted due to equipment failure or other factors, the reachability identification matrix will be updated immediately, enabling the SDN controller to quickly perceive the fault point and adjust the routing of the data stream. Compared with the traditional mine network architecture that relies on static configuration files, the SDN controller of the present invention makes the network status monitoring of the underground ring network more accurate and efficient by maintaining the reachability identification matrix in real time.

[0054] In addition to the reachability identification matrix, the data link status data structure also includes a data link performance vector, which includes key network performance parameters such as latency, bandwidth, and packet loss rate. The data transmission performance in the mine network can be affected by various environmental factors. For example, a high-dust environment may cause wireless signal attenuation, network congestion may increase transmission latency, and equipment aging may lead to a decrease in bandwidth. Therefore, relying solely on topological information cannot accurately determine the optimal path. During the process of maintaining the data link performance vector, the SDN controller of the present invention periodically collects the real-time status of each data link and dynamically evaluates the communication quality of each path by comprehensively analyzing indicators such as latency, bandwidth, and packet loss rate. When the performance of a certain data link drops below the set threshold, even if it is still in an available state, the SDN controller may preferentially switch to a backup path to ensure that the upgraded data can be transmitted to the target edge computing node with the best quality. To achieve fast fault recovery, the SDN controller not only needs to be able to perceive the network status in real time but also calculate and pre-store multiple backup paths in advance. The SDN controller of the present invention calculates and stores a self-healing backup path list periodically according to the data link status data structure. This list contains several backup paths for each main routing path, ensuring that a quick switch can be made when the network encounters an abnormality. In the traditional mine network architecture, the backup path is usually fixed and the switch is only triggered when the main route completely fails. However, the SDN controller of the present invention adopts a more intelligent backup path management strategy. That is, in the backup path list, each path entry includes a path sequence identifier, a path availability probability vector, and a path switching cost vector. The path sequence identifier is used to uniquely determine a certain backup path, and the path availability probability vector is a predicted value of the backup path availability calculated based on historical network performance data. The SDN controller can use this vector to dynamically select among multiple backup paths. The path switching cost vector is used to measure the overhead required for different path switches, including the cost of retransmitting data packets, the cost of updating the routing table, and the short-term network jitter caused by the switch. Compared with the traditional fixed-priority-based backup path switching method, the SDN controller of the present invention can make a more intelligent routing switch decision by comprehensively analyzing path availability and switching cost, ensuring the maximization of data transmission stability and reliability while minimizing the network switching overhead.

[0055] When the SDN controller calculates the list of alternative paths, it distributes these pre-calculated paths to the virtual router proxy edge computing nodes it manages, enabling each edge computing node to pre-store possible alternative routes and quickly adjust its routing table when the SDN controller triggers a path switching instruction. In a traditional mine network, path switching usually requires the centralized controller to recalculate the path and distribute new configurations, which may cause delays of several seconds or even longer. However, with the present invention, by locally storing alternative paths in the edge computing nodes, path switching can be completed within milliseconds, greatly enhancing the response speed of network fault recovery. During the upgrade process of the mine intelligent control system, if the upgrade data transmission is interrupted due to a network fault, it may cause the system upgrade to fail or the functional modules to malfunction. The SDN controller of the present invention can complete path switching instantly when a network fault occurs, ensuring the uninterrupted transmission of upgrade data and thus guaranteeing the smooth progress of system upgrade.

[0056] Embodiment 5: The virtual router proxy edge computing node maintains a local path switching forwarding table based on the list of alternative paths. The local path switching forwarding table contains a shared self-healing vector orthogonal to the primary path to dynamically mark the real-time status of each alternative path. The shared self-healing vector is shared by multiple virtual router proxy edge computing nodes and is used to identify the active or pending activation status of each path in the set of alternative paths.

[0057] When partial link failures occur in the mine network, how to quickly adjust the routing of data streams to bypass the fault area and continue transmission through the optimal alternative path is the key to ensuring the continuity and stability of the mine intelligent control system upgrade process. Through the design of the virtual router proxy edge computing node in the present invention, each edge computing node can independently maintain a local path switching forwarding table without relying entirely on the centralized scheduling of the SDN controller, thus greatly improving the speed of fault recovery and reducing the problem of network switching lag caused by centralized control calculation delay. In a traditional network architecture, path switching usually relies on a centralized control method, that is, when a primary path fails, the network controller recalculates the alternative path and distributes the new routing information to all relevant devices. However, this method is not applicable in the mine environment because the communication conditions of the underground ring network are complex, and the transmission of control signals may be affected by factors such as underground equipment failures, power supply problems, and wireless signal interference, resulting in the centralized controller being unable to complete path switching in a timely manner and thus affecting the continuity of data transmission. With the present invention, by embedding a virtual router proxy in the edge computing node, the decision of path switching can be made locally without waiting for the instruction of the SDN controller, which not only shortens the response time of path switching but also improves the adaptability of the system in the event of a sudden network fault.

[0058] The path switching ability of the virtual router proxy edge computing node depends on the local path switching forwarding table, which contains all possible alternative paths and is stored orthogonally to the primary path, enabling the alternative paths to independently update their states without interfering with the primary path. To manage the alternative path states more efficiently, the present invention introduces a shared self-healing vector, which is shared by multiple virtual router proxy edge computing nodes and is used to dynamically mark the real-time states of each alternative path. The core function of the shared self-healing vector is to provide a decentralized path activation mechanism, enabling multiple edge computing nodes to collaboratively determine the optimal alternative path when a local network failure occurs, without having to wait for global instructions from the SDN controller. Compared with traditional static alternative path switching strategies, the shared self-healing vector makes the path switching process more adaptable, capable of dynamically adjusting the priorities of alternative paths according to the real-time state of the network, thereby improving the stability and reliability of data transmission. The shared self-healing vector works dynamically. It not only stores the availability states of current alternative paths but also can intelligently sort the set of alternative paths to optimize the path switching process. In a mine network environment, the transmission performance of different alternative paths may vary over time. For example, some paths may experience increased latency due to network congestion during a specific time period, while some alternative paths may be temporarily unavailable due to equipment failures. The shared self-healing vector of the present invention can monitor these path states in real time and classify the paths in the set of alternative paths into three states: "activated", "to be activated", and "unavailable". When the primary path fails, the virtual router proxy edge computing node will preferentially select an alternative path in the "activated" state for switching. If all activated paths are unavailable, it will attempt to switch to a "to be activated" path and decide whether to continue using this path based on the actual transmission effect, while "unavailable" paths will not be selected until their states are restored to an available state.

[0059] In the specific implementation process, the shared self-healing vector can be synchronized among all edge computing nodes through a distributed update mechanism. When a virtual router acting as an edge computing node detects a change in the status of a certain backup path, such as an increase in path delay or a rise in packet loss rate, it will immediately update the local shared self-healing vector and synchronize the update information to other nodes, enabling all edge computing nodes in the mine network to obtain the latest path status information. This distributed path management method avoids the control bottleneck problem that may be caused by the traditional centralized control mode, enabling the mine network to respond more quickly in the face of sudden failures. During the upgrade process of the mine intelligent control system, the stability of data transmission is crucial. For example, when upgrading the shearer control system, if upgrade data packets are lost due to a mine network failure, it may affect the automation control logic of the shearer, thus having a serious impact on mine production. However, through the shared self-healing vector mechanism of the present invention, the upgrade data can quickly find the optimal backup path for transmission in the event of the failure of the main path, thereby ensuring the stability and continuity of the upgrade process. In addition, due to the distributed nature of the shared self-healing vector, even if the SDN controller is temporarily disconnected due to a mine network failure, the edge computing nodes can still independently complete path switching based on the existing path information, avoiding the paralysis of the entire mine network due to the unavailability of the SDN controller. The traditional path switching method is usually based on fixed priority rules, that is, the backup paths are assigned priorities during presetting, and path switching is executed in a fixed order. However, this method cannot dynamically adapt to changes in the network state, which may result in the performance of the backup path being lower than that of the current failed path in some cases, and even cause more serious network congestion. The shared self-healing vector of the present invention enables the path switching selection to be more flexible through dynamic monitoring and distributed decision-making, and can be adaptively adjusted in combination with the real-time state of the current network, improving the utilization rate of the backup path and reducing the cost of path switching. In summary, through the virtual router acting as an edge computing node and the shared self-healing vector mechanism, the present invention realizes a distributed and dynamically adaptable path switching scheme, enabling the mine intelligent control system to maintain an efficient and stable upgrade data transmission ability in a complex underground environment. Compared with the traditional centralized routing management method, the path switching method of the present invention can complete network fault recovery at the millisecond level, greatly improving the availability of the mine intelligent control system, while reducing the dependence on the SDN controller, making the entire system have stronger robustness and anti-interference ability, and providing strong technical support for the online upgrade of the mine intelligent control system.

[0060] Embodiment 6: In response to the SDN controller detecting a failure of one or more data links in the primary path, the SDN controller notifies the corresponding virtual router proxy edge computing node to update the shared self-healing vector in real time; in response to the shared self-healing vector indicating that the standby path status is the active state, the virtual router proxy edge computing node dynamically executes a network service switching process, switches the network service from the primary path where the failed data link is located to the optimal standby path in the standby path list without loss, and updates the local path switching forwarding table to reflect the current path status used by the traffic flow.

[0061] Specifically, in practical applications, the mine environment is complex and changeable, and network links may suddenly fail due to factors such as power supply failures, equipment damage, signal interference, or physical breaks. If the traditional path switching method relies on static configuration or manual intervention, the system often takes a long time to resume normal communication after detecting a primary path failure, which may result in the loss of upgrade data packets and even affect the stability of the mine intelligent control system. However, after the SDN controller of the present invention detects a failure of a certain or multiple data links in the primary path, it can notify the corresponding virtual router proxy edge computing node to update the shared self-healing vector in real time, thereby ensuring that all edge computing nodes can synchronously perceive the current network state change and complete the path switching in the shortest time. The role of the shared self-healing vector in this embodiment is crucial. It is a data structure jointly maintained by all virtual router proxy edge computing nodes, which can dynamically store the status of the current standby path and perform intelligent sorting on the standby paths to ensure that the optimal standby path can be quickly found when the primary path fails. When the SDN controller detects a primary path failure and notifies the edge computing node to update the shared self-healing vector, the vector will immediately indicate the standby path status. If a certain standby path is already in the active state, it means that the path already has sufficient bandwidth, low latency, and low packet loss rate and can be immediately put into use. At this time, the virtual router proxy edge computing node will dynamically execute a network service switching process, that is, switch the upgrade data originally transmitted through the failed data link to the optimal standby path indicated by the shared self-healing vector, and complete the path reconfiguration to ensure the continuity and stability of data transmission.

[0062] The path switching process of the present invention adopts a lossless switching mechanism, that is, after a failure occurs in the main path, data will not be lost or interrupted for a long time due to path change. The key to achieving this goal lies in the way of path switching: when the virtual router proxy edge computing node performs switching, it will first establish a parallel data stream on the standby path, and only after ensuring that the data can be normally transmitted on the standby path, will it disconnect the data stream on the main path, thus avoiding data loss or a sudden increase in transmission delay. This method is different from the traditional path switching method, which usually starts looking for a new available path after the main path fails, so it will cause a short data interruption during the switching process. The method of the present invention can achieve instantaneous switching, that is, the data stream will not be affected during the path change process, thus ensuring the stable operation of the mine intelligent control system.

[0063] In the online upgrade process of the mine intelligent control system, this lossless path switching mechanism is particularly important. For example, during the upgrade of the shearer control system, the continuity of the upgrade data is directly related to the accuracy and safety of the control logic. If the upgrade data is lost due to a failed path switch, it may lead to incorrect calculation of the control algorithm, resulting in misoperation of the shearer and even equipment damage or production safety accidents. The path switching method of the present invention can immediately complete data redirection through the standby path at the moment of the main path failure, ensuring the integrity and real-time nature of the upgrade data and avoiding affecting mine production due to network problems. In addition, after the path switching is completed, the virtual router proxy edge computing node will also update the local path switching forwarding table to reflect the path status used by the current traffic flow. The update process of the path switching forwarding table is crucial, as it determines how the subsequent data stream is routed in the mine network to ensure that all edge computing nodes can correctly identify the current network topology change and avoid sending data to the failed link after the path switch. The update of the routing table in the traditional mine network usually requires unified management by the central controller, so there is a relatively large time delay. The design of the present invention enables the virtual router proxy edge computing node to autonomously update the local path switching forwarding table, thus greatly reducing the time for path adjustment and improving the flexibility and intelligence of path switching. In a wider application scenario, such as the automatic control of the mine ventilation system, the system needs to adjust the fan operation parameters in real time to maintain a reasonable air flow distribution in the mine. If the sensing data cannot be transmitted to the control center due to a network link failure, the ventilation system may not be able to respond correctly in an emergency, affecting mine safety. In the method of the present invention, when the SDN controller detects a data link failure in the ventilation system, it will immediately trigger the path switching process of the virtual router proxy edge computing node, enabling key data such as wind speed and gas concentration to continue to be transmitted through the standby path, thus ensuring the stable operation of the ventilation control system and avoiding potential safety hazards caused by network failures.

[0064] Embodiment 7: The self-healing routing process performed by the SDN controller further includes: the virtual router agent edge computing node periodically sends a data link status detection message to the SDN controller, and the data link status detection message carries a data field identifying the real-time performance of the data link on the current path; in response to the SDN controller receiving the data link status detection message, the SDN controller updates the data link performance vector and, based on the updated data link performance vector, re-calculates the path switching cost vector in the standby path list in real time; when the updated path switching cost vector exceeds a preset threshold, the SDN controller actively triggers the virtual router agent edge computing node to perform a path pre-switching action and pre-switch to a standby path with a lower cost in the standby path list.

[0065] Specifically, in the method of the present invention, the virtual router proxy edge computing node not only undertakes the functions of path switching and traffic scheduling, but also is responsible for continuously monitoring the real-time state of the data link. Specifically, each virtual router proxy edge computing node periodically sends data link status detection messages to the SDN controller. These messages contain detailed link performance parameters, such as key performance indicators such as the delay, bandwidth, and packet loss rate of the current path. The introduction of this data link status detection mechanism enables the SDN controller to obtain the global state of the mine ring network and perform real-time evaluation of the health status of all data links in the network. Compared with the traditional network architecture that relies on static network configuration or only updates the network state when a failure occurs, the method of the present invention enables the SDN controller to perform more accurate path management based on dynamic data, thereby improving the adaptability of the mine intelligent control system to network environment fluctuations. When the SDN controller receives the data link status detection message sent by the virtual router proxy edge computing node, it updates the data link performance vector. This vector contains the latest performance indicators of all data links and recalculates the path switching cost vector in the standby path list every time it is updated. The path switching cost vector is used to measure the usage cost of different standby paths, including the degree of deterioration of the communication quality of the current path, the transmission performance of the standby path, and the possible data loss or delay fluctuation during switching. In the traditional mine network, the standby path is usually predefined and only enabled after the main path completely fails. However, in the present invention, by dynamically calculating the path switching cost, the SDN controller can actively adjust the routing of the data stream without having to wait until the main path is completely unavailable before switching. If the updated path switching cost vector exceeds the preset performance threshold, for example, the delay of a certain main path suddenly increases to twice the original value, or the bandwidth drops below the minimum acceptable level, the SDN controller will actively trigger the virtual router proxy edge computing node to perform a path pre-switching action. The core idea of path pre-switching is not to wait for the main path to completely fail, but to switch to a standby path with a lower cost in the standby path list in advance when it is detected that the network quality has deteriorated to the extent that it affects data transmission, so as to ensure that the transmission quality of the data stream will not be affected by the performance degradation of the main path.

[0066] The execution process of the path pre-switching action needs to be precisely controlled to avoid network jitter or unnecessary data flow redirection caused by overly frequent path switching. In the method of the present invention, the SDN controller not only calculates the path switching cost but also analyzes the fluctuation trend of the current path state in combination with historical data. If the performance of a certain data link only drops briefly, such as a temporary increase in delay due to network congestion, the SDN controller may not immediately trigger a path switch but wait for a short time and then detect again whether it returns to normal. If the performance degradation of the data link persists and has affected the stability of data transmission, the SDN controller will immediately trigger a path pre-switch to ensure that the data flow always stays on the optimal transmission path. In the process of upgrading the mine intelligent control system, this path pre-switching strategy is particularly important. For example, when remotely upgrading the underground equipment control system, the transmission quality of the upgrade data directly determines the success rate of the upgrade. If the data transmission rate decreases due to the performance degradation of the main path, it may lead to an extended upgrade process or even a failed upgrade. The method of the present invention can select a better alternative path in advance for switching when the performance of the main path just starts to decline, so as to ensure that the upgrade data can be transmitted with the optimal quality and ensure the smooth completion of the upgrade process of the mine intelligent control system. In addition, the path pre-switching mechanism of the present invention can also effectively cope with the dynamic environmental changes of the mine network. For example, in the intelligent control scenario of the mine ventilation system, the gas concentration sensor needs to transmit data to the central control system in real time to adjust the fan operation parameters in a timely manner. If the delay of the data link suddenly increases and the control system still adjusts the fan operation state according to the old data, it may lead to uneven gas flow inside the mine and increase the safety risk. The path pre-switching mechanism of the present invention can switch the data flow to the low-delay alternative path before the delay reaches the critical value to ensure that the gas concentration data can be transmitted in real time, thus ensuring the safe operation of the ventilation system.

[0067] Embodiment 8: The dynamic execution of the network service switching process includes: the virtual router proxy edge computing node maintains a consistent hashing mapping for the currently active network service path, and the consistent hashing mapping maps the network service flow identifier to a set of edge computing nodes on the hashing ring; in response to the change of the path state from active or pending activation, the consistent hashing ring is dynamically updated to exclude the edge computing nodes or data links corresponding to the faulty data link, and the transmission path of the network service flow is remapped to ensure the uniform distribution and load balancing of the network service flow; during the service switching process, the continuity of the original service flow and the hashing ring mapping is maintained to avoid service jitter caused by frequent path switching in a short period of time.

[0068] Specifically, in the solution of the present invention, the virtual router proxy edge computing node adopts consistent hashing mapping to maintain the currently active network service path. The core function of this mechanism is to ensure that different data streams can still be transmitted along the optimal path when the network topology changes, and can avoid sudden interruption of service flows or load imbalance caused by path adjustment. The basic principle of consistent hashing mapping is to map the identifiers of network service flows (such as IP addresses, port numbers, or flow IDs) to a hash ring, and allocate them to different edge computing nodes according to the hash values. In the mine intelligent control system, the data streams may include mine monitoring data, equipment control instructions, remote upgrade data, etc. The reliability of these service flows is directly related to the safety of mine production. Therefore, when the path is adjusted, it is crucial to ensure the uniform distribution and continuity of the service flows. Specifically, in the solution of the present invention, the virtual router proxy edge computing node adopts consistent hashing mapping to maintain the currently active network service path. The core function of this mechanism is to ensure that different data streams can still be transmitted along the optimal path when the network topology changes, and can avoid sudden interruption of service flows or load imbalance caused by path adjustment. The basic principle of consistent hashing mapping is to map the identifiers of network service flows (such as IP addresses, port numbers, or flow IDs) to a hash ring, and allocate them to different edge computing nodes according to the hash values. In the mine intelligent control system, the data streams may include mine monitoring data, equipment control instructions, remote upgrade data, etc. The reliability of these service flows is directly related to the safety of mine production. Therefore, when the path is adjusted, it is crucial to ensure the uniform distribution and continuity of the service flows.

[0069] In addition, during the remote upgrade process, the transmission path of the upgrade data needs to be highly reliable because any packet loss may lead to upgrade failure or abnormal device functions. If the transmission quality of a certain data link deteriorates during the upgrade process, traditional path switching methods usually immediately switch the entire data stream to a new path, which may cause drastic changes in the transmission paths of all upgrade data in a short period of time, thereby triggering unnecessary network fluctuations. In the method of the present invention, the virtual router proxy edge computing node will, based on the adjustment strategy of hash mapping, only redirect the affected data stream without affecting the data transmission of the entire network, thus ensuring the stability of the upgrade process. Another key advantage of consistent hash mapping lies in load balancing, that is, when adjusting the path, instead of simply finding an available path, based on the hash ring mechanism, the new data stream can be evenly distributed among all available edge computing nodes, thus avoiding the problem of traffic concentration on a few paths resulting in local congestion. For example, during the upgrade process of the mine production optimization scheduling system, the system needs to continuously receive production data and calculate the optimal scheduling strategy. If a certain computing node is overloaded due to path switching, it may affect the timeliness of scheduling decisions. Through consistent hash mapping, the present invention can ensure load balancing of the service flow after path adjustment, making the task volume of each computing node evenly distributed, thereby improving the overall computing efficiency of the system. In the method of the present invention, in order to further improve the smoothness of path adjustment, the virtual router proxy edge computing node also adopts the principle of minimum perturbation, that is, when adjusting the path, as much as possible, the mapping relationship of the original data stream remains unchanged, and only the smallest affected subset is adjusted. The application of this principle enables the mine intelligent control system to minimize the transmission fluctuations of upgrade data or control instructions when the network changes. For example, during the upgrade process of the mine power supply monitoring system, if some transmission data streams need to be reallocated due to path adjustment, and these data are related to real-time power load monitoring, any transmission instability may affect power supply safety. The method of the present invention can ensure that even when the network is adjusted, the original service flow can still be transmitted along the optimal path, thereby ensuring the stable operation of mine production.

[0070] Embodiment 9: The delay and packet loss rate in the data link performance vector are processed by the SDN controller using a statistical sliding window algorithm, and the least squares method is used to predict the evolution trend of the data link performance for predicting the probability of data link failure occurrence; when the predicted evolution trend of the data link performance exceeds the set warning threshold, the SDN controller actively issues a preventive standby path switching instruction to the virtual router proxy edge computing node in advance.

[0071] Specifically, the stability of the mine intelligent control system highly depends on the reliability of the underground network. The traditional network management method is usually based on passive fault detection, that is, path adjustment is only carried out after the data link completely fails or its performance severely degrades. However, in the complex network environment of the mine, the degradation of the data link is often a gradual process. For example, factors such as equipment aging, signal interference, and network congestion can lead to increased latency or elevated packet loss rate. If only relying on the emergency handling method after a fault occurs, it may not only cause the interruption of the upgrade data transmission of the mine intelligent control system, but also affect the stability of mine production scheduling and safety monitoring. Therefore, the present invention proposes an active prediction and early optimization path switching method, enabling the SDN controller to take preventive measures before a fault actually occurs to ensure the high reliability of the data link. In the method of the present invention, the latency and packet loss rate in the data link performance vector are processed by the SDN controller using the statistical sliding window algorithm. The core idea of the statistical sliding window algorithm is to select a time window from the historical performance data of the data link and calculate key statistical indicators such as the mean, variance, and extreme values of the data within this window to observe the short-term performance fluctuations of the data link. Compared with the traditional single-point measurement method, the sliding window method can effectively smooth out sudden abnormal fluctuations and avoid the impact of short-term jitters on the network state assessment. For example, in the mine wireless communication link, due to the movement of underground vehicles or the operation of fans, instantaneous signal interference may occur, but these short-term fluctuations usually do not affect the long-term data transmission performance. Therefore, using the sliding window to calculate the trend of network performance can more accurately capture the true evolution trend of the data link. However, simply relying on the sliding window to calculate the current state of network performance is still insufficient for accurate prediction. Therefore, the present invention further uses the least squares method to predict the evolution trend of the data link performance. The least squares method is a regression analysis method that can fit the best trend line based on the historical performance data of the data link to predict the state of the data link in the future period. Specifically, the SDN controller will use the data within the sliding window to calculate the change trends of latency and packet loss rate, and fit a time evolution model through the least squares method to predict the future state of the data link. For example, if the packet loss rate of a certain data link shows an exponential growth trend in the past 30 minutes, it can be inferred that the link is likely to experience severe degradation or even complete failure in the next 10 - 20 minutes. In the traditional mine network management method, path switching is usually only triggered after large-scale packet loss or complete interruption occurs in the data link, while the prediction method of the present invention can identify potential risks in advance before the network performance deteriorates to the critical point and actively take optimization measures.

[0072] When the predicted evolution trend of the data link performance exceeds the set warning threshold, the SDN controller will actively issue a preventive standby path switching instruction to the virtual router proxy edge computing nodes in advance. Here, the warning threshold is not a fixed value but a dynamically adjusted threshold based on self-adaptation. Specifically, the SDN controller will dynamically adjust the threshold for triggering path switching by combining historical data, the current network state, and service requirements. For example, during the peak production period in a mine, the network load is high, and higher reliability is required for the data link. Therefore, the warning threshold can be appropriately lowered to trigger path switching earlier. During the low-load period, the network has stronger fault tolerance, and the threshold can be appropriately increased to reduce unnecessary path switching and prevent network jitter. The role of the preventive path switching instruction is to migrate the data stream to the standby path before the data link fails completely, avoiding network interruption caused by sudden failures. For example, in a mine gas monitoring system, if the delay trend of a certain data link continues to rise without taking measures, it may result in the inability to transmit gas concentration data to the control center in real time, thus affecting the timeliness of safety warnings. The method of the present invention can switch the data stream in advance when it detects that the performance of the link continues to decline, enabling it to continue to be transmitted through the standby path, thereby ensuring the stable operation of the gas monitoring system. Similarly, during the upgrade process of a mine intelligent control system, if the upgrade data is lost due to network problems, it may lead to equipment upgrade failure. The preventive path switching method of the present invention can ensure the transmission of upgrade data under the best network conditions, greatly improving the success rate of the upgrade. In addition, the preventive path switching instruction of the present invention does not take effect immediately but adopts a progressive switching strategy. That is, when the data link state has not completely failed, the data stream weight of the standby path is gradually increased to make the migration process of the data stream smoother. For example, in a mine production optimization system, if the delay of a data link is slowly rising, the SDN controller will not directly switch all data streams but first guide some low-priority data streams to the standby path and observe the change of network performance. If the performance of the main path still continues to decline, then gradually migrate the high-priority data streams to the standby path to ensure the stability of data transmission. This progressive switching strategy can minimize service jitter to the greatest extent compared with the traditional sudden path switching method, ensuring the stable operation of the mine intelligent control system.

[0073] Embodiment 10: Among them, self-healing vectors are shared in real time among multiple virtual router proxy edge computing nodes through a multicast channel, and a message queue and a distributed consistency protocol are used to ensure the consistency of the path switching state; among them, the path switching cost vector further includes the energy consumption index of the path, and the SDN controller preferentially selects a standby path with lower energy consumption for switching based on path energy consumption prediction to achieve the goal of network green energy conservation.

[0074] In the method of the present invention, multiple virtual router proxy edge computing nodes share self-healing vectors in real time through a multicast channel. The introduction of this mechanism enables the path status information in the mine network to be quickly synchronized among all edge computing nodes, ensuring that all path switching decisions are based on the latest network status information. Traditional network architectures usually adopt a point-to-point path update method. When the status of a certain path changes, all relevant devices need to be notified one by one. However, this method is vulnerable to network latency and communication instability in the mine environment, resulting in asynchronous path switching information among different nodes. The present invention uses a multicast channel, enabling all virtual router proxy edge computing nodes to receive and update the shared self-healing vector simultaneously, thereby ensuring the synchronization and consistency of the path status. In addition, to ensure the consistency of the path switching status among all edge computing nodes, the present invention adopts a message queue and a distributed consistency protocol to ensure that there will be no inconsistent data flow redirection during the path switching process. In a mine intelligent control system, different edge computing nodes may be located in different areas of the mine. If the synchronization of path switching information cannot be guaranteed, it may lead to some nodes still using a failed path while some nodes have completed the path switching, resulting in unstable data transmission. For example, in a mine gas monitoring system, if an edge computing node erroneously believes that the main path is still available while another node has switched to a backup path, it may result in the loss or repeated transmission of sensing data, affecting the accuracy of gas concentration monitoring. The message queue mechanism of the present invention ensures that all path switching instructions are executed in a strict chronological order, and all edge computing nodes can make consistent path switching decisions based on the same path status data. The distributed consistency protocol (such as RAFT or Paxos) further ensures the coordination of the path switching status among multiple edge computing nodes, enabling the consistency of path switching to be guaranteed even in the case of network environment fluctuations. In addition to the consistency of path switching, the present invention also adds the energy consumption index of the path to the path switching cost vector, and the SDN controller preferentially selects a backup path with lower energy consumption for switching based on path energy consumption prediction to achieve the goal of green energy conservation in the network. Traditional path switching algorithms usually only consider the delay, bandwidth, and packet loss rate of the path. However, in the mine network environment, some paths may need to rely on high-power consumption devices (such as signal amplifiers, relay stations, etc.) to maintain stable communication, while some other paths may be natural low-power consumption paths (such as fiber optic direct connection links, short-distance wireless communication links, etc.). If the energy consumption factor is not considered during the path switching process, it may lead to too high energy consumption overhead in the mine intelligent control system, which does not conform to the sustainable development goal of green mines.Therefore, the path energy consumption optimization strategy of the present invention enables the SDN controller to not only consider communication quality when selecting an alternative path, but also comprehensively evaluate the energy consumption of different paths, and preferentially use low-power links for data transmission, thereby reducing the overall energy consumption while ensuring communication stability.

[0075] During the process of path energy consumption optimization, the SDN controller models the energy consumption characteristics of different data links, and based on historical data and the current state, predicts the energy consumption change trends of different paths in the future. For example, in a mine network, some wireless communication links have high energy consumption under high load conditions, while the energy consumption of fiber optic links is relatively stable. Therefore, under low load conditions, the SDN controller may preferentially use wireless links to improve flexibility, and under high load conditions, it will preferentially switch to fiber optic links to reduce energy consumption. In addition, the energy consumption optimization algorithm of the present invention also comprehensively considers the cost of path switching. For example, if the switching cost of a low-power path is high (such as requiring a long time for signal resynchronization), the SDN controller may choose a path with slightly higher energy consumption but lower switching cost to ensure energy conservation without affecting the continuity of data transmission. In an actual application scenario, for example, during the online upgrade of a mine intelligent ventilation control system, if the control instruction transmission link of the system needs to be switched due to a fault without considering energy consumption issues, it may cause the fan control signal to be transmitted through a high-power path, thereby increasing the overall energy consumption of the mine. The method of the present invention can ensure that the transmission path of the ventilation control instruction preferentially selects a low-power path on the premise of ensuring delay and reliability, thereby realizing the intelligence and energy-saving optimization of the fan control system. In addition, during the upgrade of the mine video surveillance system, due to the high energy consumption of high-definition video data transmission, the SDN controller of the present invention can dynamically calculate the energy consumption of different alternative paths during path switching and select the path with the optimal energy consumption and sufficient bandwidth to ensure the efficient transmission of video data while reducing the overall power consumption of the mine intelligent monitoring system.

[0076] The preferred embodiments of the present disclosure have been described above with reference to the accompanying drawings, and the scope of the rights of the present disclosure is not limited thereby. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present disclosure shall fall within the scope of the rights of the present disclosure.

Claims

1. A method for upgrading a mine intelligent control system supported by an underground ring network, characterized in that, The method includes: Step 1: Deploy edge computing nodes underground. Each edge computing node corresponds to an independent functional module of the mine intelligent control system; each edge computing node is connected to the underground ring network; each edge computing node includes: a twin part and an upgrade part; the twin part is a real-time mirror of the functional module corresponding to this edge computing node. Step 2: When an upgrade is required for a certain functional module, the upgrade data is sent to the underground ring network. The underground ring network controls the upgrade data through the SDN controller according to the header in the upgrade data, and through routing control, transmits it to the edge computing node corresponding to this functional module via the data link. Step 3: After receiving the upgrade data, the upgrade part of the edge computing node isolates the functional module from the mine intelligent control system, and the twin part replaces the functional module to realize the corresponding functions in the mine intelligent control system; the upgrade part transmits the upgrade data to the isolated functional module for upgrading; at the beginning of the upgrade, the edge computing node first isolates the functional module to be upgraded, that is, strips the real-time control authority of this module from the mine intelligent control system to ensure that other system functions will not be affected during the upgrade process; the key to the isolation process is to ensure the integrity and consistency of the data. When isolation occurs, the edge computing node needs to ensure that the twin part has completely synchronized the running state of the current functional module and can independently execute all calculation tasks of this module, so that the entire mine intelligent control system can not perceive the switching of the functional module, thereby avoiding affecting the continuity of mine production due to instability during the switching process. Step 4: After the upgrade is completed, the upgraded part isolates the twin part from the mine intelligent control system and lifts the isolation of the functional module, allowing the functional module to reconnect to the mine intelligent control system; before the twin part exits, a system status comparison and consistency verification will be carried out first to ensure that the upgraded functional module is consistent with the operation results of the twin part in terms of data input, logic processing, control instruction execution, etc.; during the process where the twin part temporarily replaces the functional module to execute tasks, it has accumulated a certain amount of system data, and this system data will be used as the benchmark for status comparison. Before taking over, the upgraded functional module needs to pass through this comparison process to ensure that all key variables, system parameters, and control states can be correctly docked, thus avoiding system anomalies caused by status drift or data mutation; after the status comparison is completed, the upgraded part issues a switching instruction to gradually transfer the control right of the twin part to the upgraded functional module; this process is a progressive dynamic adjustment process carried out under the condition of a stable operation state of the mine intelligent control system to avoid the impact of sudden switching on the entire mine intelligent control system; by adopting the progressive dynamic switching method, the twin part and the upgraded functional module are simultaneously operated within a short period of time, enabling the control signals, data streams, and calculation logics of the mine intelligent control system to smoothly transition between the two; during the upgrade period, there may be differences between the twin part and the upgraded functional module in terms of data storage, log recording, and status caching. Therefore, before the twin part exits, data synchronization is required to ensure that all relevant historical data can be correctly transmitted to the upgraded functional module and stored and updated internally.

2. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 1, characterized in that, The functional modules of the mine intelligent control system at least include: a monitoring and sensing module for real-time monitoring of the mine environment, equipment operation status, and personnel safety, a data acquisition and communication module for underground data acquisition, transmission, and storage, an intelligent control and automation module for automatic control of various mine equipment, and a production optimization and intelligent scheduling module for scheduling and optimizing the mine production process.

3. The method for upgrading the mine intelligent control system supported by the underground ring network according to claim 2, wherein, When a data link or route fails due to equipment failure or power supply interruption in the underground environment of the underground ring network, the SDN controller executes a self-repair routing process to ensure that the upgrade data is sent to the corresponding functional module via other data links under the control of the routing. The routing includes: a main route and a backup route.

4. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 3, characterized in that, The SDN controller executes the self-healing routing process, which specifically includes: the SDN controller maintains the global topology information and data link status data structure for multiple edge computing nodes. The data link status data structure includes the edge computing node reachability identification matrix and the data link performance vector between edge computing nodes. The edge computing node reachability identification matrix orthogonally identifies the current active state of each data link or edge computing node for the paths between multiple edge computing nodes. The data link performance vector includes one or more of delay, bandwidth, and packet loss rate; the SDN controller periodically calculates and pre-stores the self-healing alternative path list according to the data link status data structure. The alternative path list contains several alternative routes for each primary routing path, and each path entry in the alternative path list includes a path sequence identifier, a path availability probability vector, and a path switching cost vector; the SDN controller distributes the pre-calculated alternative path list to the virtual router proxy edge computing nodes it manages.

5. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 4, characterized in that, The virtual router proxy edge computing nodes maintain a local path switching forwarding table based on the alternative path list. The local path switching forwarding table contains a shared self-healing vector orthogonal to the primary path to dynamically mark the real-time state of each alternative path. The shared self-healing vector is shared by multiple virtual router proxy edge computing nodes and is used to identify the active or to-be-activated state of each path in the alternative path set.

6. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 5, characterized in that, In response to the SDN controller detecting a failure of one or more data links in the primary path, the SDN controller notifies the corresponding virtual router proxy edge computing nodes to update the shared self-healing vector in real time; in response to the shared self-healing vector identifying the alternative path state as the active state, the virtual router proxy edge computing nodes dynamically execute the network service switching process, switch the network service from the primary path where the failed data link is located to the optimal alternative path in the alternative path list without loss, and update the local path switching forwarding table to reflect the current path state used by the traffic flow.

7. The method for upgrading the mine intelligent control system supported by the underground ring network according to claim 6, characterized in that, The SDN controller executing the self-healing routing process further includes: the virtual router proxy edge computing nodes periodically send data link status probe messages to the SDN controller. The data link status probe messages carry data fields identifying the real-time performance of the data links on the current path; in response to the SDN controller receiving the data link status probe messages, the SDN controller updates the data link performance vector and, according to the updated data link performance vector, re-calculates the path switching cost vector in the alternative path list in real time; when the updated path switching cost vector exceeds the preset threshold, the SDN controller actively triggers the virtual router proxy edge computing nodes to execute the path pre-switching action and pre-switch to an alternative path with a lower cost in the alternative path list.

8. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 7, characterized in that, The dynamic execution of network service switching process includes: the virtual router proxy edge computing node maintains a consistent hashing mapping for the currently active network service path, and the consistent hashing mapping maps the network service flow identifier to the set of edge computing nodes on the hashing ring; in response to the activation of the path state or the change of the to-be-activated state, the consistent hashing ring is dynamically updated to exclude the edge computing nodes or data links corresponding to the faulty data links, and the transmission path of the network service flow is remapped to ensure the uniform distribution and load balancing of the network service flow; during the service switching process, the continuity of the original service flow and the hashing ring mapping is maintained to avoid service jitter caused by frequent path switching in a short period of time.

9. The method for upgrading the intelligent control system of a mine supported by an underground ring network according to claim 8, characterized in that, Among them, the delay and packet loss rate in the data link performance vector are processed by the SDN controller using the statistical sliding window algorithm, and the least square method is used to predict the evolution trend of the data link performance, which is used to predict the probability of data link failure; when the predicted evolution trend of the data link performance exceeds the set warning threshold, the SDN controller actively issues a preventive standby path switching instruction to the virtual router proxy edge computing node in advance.

10. The method for upgrading the mine intelligent control system supported by the downhole ring network according to claim 9, characterized in that, Among them, multiple virtual router proxy edge computing nodes share the self-healing vector in real time through the multicast channel, and use the message queue and distributed consistency protocol to ensure the consistency of the path switching state; among them, the path switching cost vector also includes the energy consumption index of the path, and the SDN controller preferentially selects a standby path with lower energy consumption for switching based on the path energy consumption prediction to achieve the goal of network green energy saving.

Images