Shaft lifting system and method based on intelligent control

Abstract

The invention relates to the technical field of shaft lifting system control, and discloses a shaft lifting system and method based on intelligent control. According to the system, an acquisition module collects an operation table signal and an acousto-optic annunciator state to obtain an operation instruction and equipment state data; the marking module adds a timestamp and constructs an intrinsic safety type distributed control topological structure; the input module sends the state data to the sensor network to identify the lifting state; the implementation module establishes a dependency graph to realize a horizontal forwarding locking function; the matching module performs feature extraction and fault identification; and the association module analyzes and processes the strategy and the process record, and outputs a parameter optimization value and a locking logic adjustment scheme. According to the application, the distributed control network is constructed, state analysis is realized in combination with the node link sensor network, and continuous improvement of system performance is realized through data-driven fault diagnosis and parameter optimization, so that the safety, reliability and operation efficiency of the mine hoisting system are remarkably improved.

Description

Technical Field

[0001] The present application relates to the technical field of shaft hoisting system control, and in particular to a shaft hoisting system and method based on intelligent control. Background Art

[0002] The mine shaft hoisting system is a key equipment in mine production, responsible for the transportation of personnel, materials and equipment between the mine and underground. The traditional mine hoisting signal system mainly adopts mechanical or simple electrical structure, and realizes communication between the upper and lower wellheads through mechanical devices such as buttons and pull wires or sound and light signal devices. With the development of technology, some mines have introduced automatic hoisting control systems based on PLC to realize basic signal transmission and safety interlocking functions. In the prior art, mine signal transmission equipment such as B66B type equipment can realize automatic conversion or moving sidewalks as life-saving devices to replace normal exits; A62B1 / 02 equipment can be used in conjunction with gas engines or installed on aircraft to indicate rescue materials or devices for passengers to escape; B64D9 / 00 equipment is used to lift or hold aircraft components to unload or stop loading. These devices play an important role in their respective fields and provide basic guarantees for mine safety production.

[0003] However, the existing mine hoisting signal system has many deficiencies. First, traditional signal systems generally lack intelligent fault diagnosis and processing capabilities. Once a fault occurs, manual troubleshooting is often required, which is time-consuming, labor-intensive and prone to misjudgment. Second, the signal transmission mechanism of the existing system is relatively rigid and lacks adaptive adjustment capabilities. Signal conflicts or losses are prone to occur under complex working conditions. Third, the horizontal forwarding interlocking function is not fully implemented, and it is unable to effectively handle the signal coordination problem between multiple levels, affecting the hoisting efficiency and safety. Fourth, the system parameter optimization mainly relies on experience settings, lacks self-optimization capabilities based on data analysis, and is difficult to adapt to the dynamic changes in the mine environment. Most importantly, most of the existing systems use isolated control methods, which makes it difficult to achieve collaborative work between control units, and cannot form a complete closed-loop intelligent control process. The reliability and adaptability in complex mine environments are limited. Summary of the invention

[0004] The present application provides a shaft hoisting system and method based on intelligent control, which is used to build a distributed control network, combine the node link sensor network to realize state analysis, apply the horizontal forwarding locking mechanism to ensure the security of signal transmission, and realize continuous improvement of system performance through data-driven fault diagnosis and parameter optimization, thereby significantly improving the safety, reliability and operation efficiency of the mine hoisting system.

[0005] In a first aspect, the present application provides a shaft hoisting system based on intelligent control, the shaft hoisting system based on intelligent control comprising: The acquisition module is used to collect the operation console signals and the status information of the sound and light signal devices at each operation point in the mine shaft hoisting system to obtain the hoisting operation instruction data and equipment status data; A marking module, used for timestamping the lifting operation instruction data, constructing a field control network through a controller, and obtaining an intrinsically safe distributed control topology structure; An input module, used for inputting the equipment status data into a lifting status analysis model based on a node-linked sensor network, identifying the lifting status, and obtaining a shaft lifting control instruction; An implementation module, used to establish a trigger dependency diagram between execution units according to the shaft hoisting control instruction, implement the hoisting signal linkage control with horizontal forwarding locking function, and obtain real-time execution feedback data and hoisting process records; A matching module, used to perform feature extraction and pattern matching on the real-time execution feedback data, identify and locate abnormal states based on a preset fault feature library, and obtain a hierarchical processing strategy; The correlation module is used to correlate and analyze the hierarchical processing strategy and the lifting process record to obtain the control parameter optimization value and the safety locking logic adjustment plan.

[0006] In a second aspect, the present application provides a shaft hoisting method based on intelligent control, the shaft hoisting method based on intelligent control comprising: Collect the operation console signals and sound and light signal status information of each operation point in the mine shaft hoisting system to obtain the hoisting operation instruction data and equipment status data; The lifting operation instruction data is timestamped, and a field control network is constructed through a controller to obtain an intrinsically safe distributed control topology structure; Inputting the equipment status data into a lifting status analysis model based on a node-linked sensor network to identify the lifting status and obtain a shaft lifting control instruction; According to the shaft hoisting control instruction, a trigger dependency diagram between execution units is established, and a hoisting signal linkage control with a horizontal forwarding locking function is implemented to obtain real-time execution feedback data and hoisting process records; Perform feature extraction and pattern matching on the real-time execution feedback data, identify and locate abnormal states based on a preset fault feature library, and obtain a hierarchical processing strategy; The hierarchical processing strategy and the lifting process record are correlated and analyzed to obtain the optimized value of the control parameter and the safety locking logic adjustment plan.

[0007] In a third aspect, a shaft hoisting device based on intelligent control is provided, comprising: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the shaft hoisting device based on intelligent control executes the above-mentioned shaft hoisting method based on intelligent control.

[0008] In a fourth aspect, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium is run on a computer, the computer is enabled to execute the above-mentioned shaft hoisting method based on intelligent control.

[0009] In the technical solution provided by the present application, by collecting signal data from each operating point of the mine shaft hoisting system, multi-level collaborative control is achieved, which significantly improves the safety, reliability and operating efficiency of the hoisting system; the hoisting operation instruction data is timestamped and an intrinsically safe distributed control topology is constructed, which solves the problems of unstable signal transmission and poor adaptability to the harsh underground environment in traditional systems, and ensures accurate timing coordination between execution units; the hoisting state analysis model based on the node-linked sensor network can accurately identify the five hoisting states of fast up, fast down, slow up, slow down and emergency stop, and converts uncertain sensor data into clear control instructions, reflecting the technical contribution of artificial intelligence algorithms in specific application fields, especially through the three-layer neural network structure to achieve high-precision state recognition, providing reliable guarantee for hoisting control in complex mine environments; the implementation of the dependency graph and the horizontal forwarding locking function effectively solves the signal conflict problem in the multi-level hoisting system, especially the problem that the lower shaft mouth can only move upward. The one-way control logic of the wellhead signal simplifies the system operation mechanism and improves safety. The feature extraction and fault matching functions of the real-time execution feedback data realize the early detection and precise positioning of faults, reducing the risk of equipment damage and downtime. The formulation of the graded processing strategy takes into account both safety and operational efficiency, and adopts differentiated processing measures according to the fault risk level. The control parameter optimization value and safety interlock logic adjustment scheme obtained through correlation analysis enable the system to have self-optimization capabilities, and can continuously adjust the working parameters and control logic according to the operating data to adapt to the challenges brought by changes in the mine environment and equipment aging. It is particularly worth emphasizing that the artificial intelligence algorithm in this scheme not only improves the accuracy of fault diagnosis through feature matching, but also clarifies the direction of improvement through sensitivity analysis in the parameter optimization process, reflecting the targeted contribution of the algorithm model in specific application scenarios, transforming traditional empirical judgments into data-driven decisions, and significantly improving the intelligence level and operation effect of the mine hoisting system. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0011] Figure 1 A schematic diagram of an embodiment of a shaft hoisting system based on intelligent control in an embodiment of the present application; Figure 2 A schematic diagram of an embodiment of a shaft hoisting method based on intelligent control in an embodiment of the present application; Figure 3 It is a schematic block diagram of the structure of a shaft hoisting device based on intelligent control in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] The embodiment of the present application provides a shaft hoisting system and method based on intelligent control. The terms first, second, third, fourth, etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms include or have and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0013] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of the shaft hoisting system based on intelligent control includes: The acquisition module 101 is used to collect the operation console signals and the state information of the sound and light signal devices at each operation point in the mine shaft hoisting system to obtain the hoisting operation instruction data and the equipment state data; The marking module 102 is used to timestamp the lifting operation instruction data, build a field control network through the controller, and obtain an intrinsically safe distributed control topology structure; An input module 103 is used to input the equipment status data into a lifting status analysis model based on a node-linked sensor network, identify the lifting status, and obtain a shaft lifting control instruction; An implementation module 104 is used to establish a trigger dependency diagram between execution units according to the shaft hoisting control instruction, implement the hoisting signal linkage control with horizontal forwarding locking function, and obtain real-time execution feedback data and hoisting process records; The matching module 105 is used to extract features and perform pattern matching on the real-time execution feedback data, identify and locate abnormal states based on a preset fault feature library, and obtain a hierarchical processing strategy; The association module 106 is used to associate and analyze the hierarchical processing strategy and the lifting process record to obtain the control parameter optimization value and the safety lockout logic adjustment plan.

[0014] It is understandable that the execution subject of the present application may be a shaft hoisting system based on intelligent control, or a terminal or a server, which is not specifically limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0015] In the embodiment of the present application, the acquisition module 101 collects the signal of the lifting system by installing a TH12 (C) type operating table and a KXH15 (B) type sound and light signal device at the upper wellhead, middle level and lower wellhead of the vertical shaft. These devices collect raw data through the S7-1200 controller, and then perform Butterworth low-pass filtering on these data, with the cutoff frequency set to 2Hz, which effectively eliminates the high-frequency interference generated by the underground electrical equipment. The processed signal is encoded according to the signal source identification, signal type identification, signal value, timestamp and status flag to form a standardized information packet, and then the data integrity is ensured by CRC32 verification.

[0016] After receiving the standardized information packet, the marking module 102 uses the high-precision time protocol (PTP) to add a time mark accurate to the millisecond level to each data. These timestamped instruction data are then input into the controller network and organized according to a three-layer network architecture, namely the bottom field layer (operating console, annunciator and other execution equipment), the middle control layer (S7-1200 and ST40 controllers) and the upper management layer (KXT131 (A) -X machine room display device). A dual redundant communication link is established through PROFIBUS-DP and PROFINET buses, and the improved Raft distributed consensus algorithm is applied to achieve state consistency between nodes. At the same time, the health status of the node is monitored through the heartbeat mechanism. The heartbeat cycle is set to 500 milliseconds. If there is no response for three consecutive times, the node is judged to be offline. Construct an intrinsically safe distributed control topology that meets the requirements of the "Coal Mine Safety Regulations".

[0017] The input module 103 inputs the equipment status data into the node link sensor network for status analysis. The network first extracts the time series features of the data, processes the continuous data frames using a 256-point sliding window, and calculates a 64-dimensional feature vector. These feature vectors are input into a three-layer network, including 64 input neurons, 128 hidden layer neurons (using the ReLU activation function) and 5 output neurons (using the Softmax function), and output five lifting states. In the embodiment of the present application, the five lifting states include: fast up, fast down, slow up, slow up and emergency stop probability distribution. These probability distributions are processed through the Markov decision process model, considering the state transition constraints, and combined with the knowledge graph for matching to generate shaft lifting control instructions.

[0018] The implementation module 104 parses the control instructions into a sequence of execution actions and constructs a dependency model in the form of a directed acyclic graph. In view of the characteristics of the mine hoisting system, a horizontal forwarding interlocking function is implemented to ensure that the lower wellhead can only send signals to the upper wellhead. After the upper wellhead receives the signal from the lower wellhead, the horizontal interlocking is completed. The execution process is divided into three stages: preparation, execution, and completion, and a 5ms status check cycle is set. A double confirmation mechanism is adopted, requiring two confirmations for instruction reception and execution completion, and a 25ms confirmation timeout threshold is set. The system collects the response data of the execution unit at a frequency of 10Hz, records the timestamp, device ID, operation type, and execution results, and stores these data through a three-level database structure (local cache, regional storage, and central warehouse), and sets a 7-day rolling update strategy.

[0019] The matching module 105 extracts mining features from the real-time execution feedback data, calculates the feedback signal of the sound and light signal device, the status data of the operating console and the response time of the execution unit. These features are compared with the preset mining fault feature library to identify specific types such as sound and light signal fault, operating console fault, transmission line fault and controller fault. The system accurately locates the level and equipment where the fault occurs, evaluates the impact of the fault on the lifting operation, and determines whether it affects signal transmission, blocks horizontal forwarding or causes the emergency stop signal to be triggered incorrectly. Based on this, the fault is divided into three levels: low, medium and high, and corresponding processing strategies are formulated.

[0020] The association module 106 performs an association analysis on the hierarchical processing strategy and the lifting process record. First, the fault handling effect is statistically analyzed to calculate the repair time, recovery success rate and recurrence frequency of different fault types. These data are associated with the lifting process record in time series to match the change trend of system operation parameters before and after the fault, and determine the key parameters affecting the system stability. The influence weight of each parameter on the system stability is determined through sensitivity analysis, and the optimal parameter value is calculated. At the same time, the triggering and release conditions of the lockout event are analyzed, and the signal transmission priority, lockout triggering conditions and lockout release timing of the upper and lower wellheads and intermediate levels are optimized to form a complete safety lockout logic adjustment plan.

[0021] In a specific embodiment, the acquisition module 101 is used to: An operating table and an acoustic and optical signal device are installed at the upper wellhead, middle level and lower wellhead of the shaft respectively. Signals are collected through the controller to obtain the original data including lifting instructions and equipment status. The original data is filtered through a Butterworth low-pass filter to obtain a boost operation signal; Digitally encode the lifting operation signal according to the signal source identifier, signal type identifier, signal value, time stamp and status flag to obtain a standardized information packet; Perform CRC32 check and encryption on the standardized information packets, and transmit them to the intelligent processing unit via industrial Ethernet to obtain the system status data; A digital twin model of the signal collection point is established based on the lifting system status data, and the virtual mapping structure of the shaft lifting system is obtained by updating the status parameters in real time; The virtual mapping structure is compared and analyzed with the historical collection data, and abnormal collection points are identified through deviation calculation to obtain improved operation instruction data and equipment status data.

[0022] Specifically, the acquisition module 101 deploys data acquisition devices at key locations in the mine shaft, including installing TH12 (C) type operating consoles and KXH15 (B) type sound and light signalers at the upper wellhead, middle level and lower wellhead respectively. The TH12 (C) type operating console is an intrinsically safe operating device for mining, which has explosion-proof characteristics and can work stably in the harsh environment of damp and dusty underground; while the KXH15 (B) type sound and light signaler has both sound prompt and light signal prompt functions, which is convenient for transmitting information in noisy environments and low light conditions. These devices collect signals through the S7-1200 programmable controller, and the controller collects data every 100 milliseconds to form a raw data stream. The raw data contains lifting instructions and equipment status.

[0023] The collected raw data is subjected to noise filtering. There are a large number of electrical equipment in the mine environment, and the electromagnetic interference generated can cause signal fluctuations. As a linear filter, the Butterworth low-pass filter has a flat amplitude-frequency characteristic in the passband and a steep cut-off characteristic, which is particularly suitable for processing noise signals in the mine environment. In this solution, a 4th-order Butterworth low-pass filter is used, and the cut-off frequency is set to 2Hz, which can effectively filter out high-frequency noise while retaining the main characteristics of the original signal. The signal after filtering is smoother, reducing the probability of false triggering.

[0024] The lifting operation signal after filtering needs to be standardized and encoded. The standardization process converts the signal into a unified format data packet according to the preset format, which contains five key fields: signal source identification (16 bits, used to indicate the device ID of the data source), signal type identification (8 bits, distinguishing between lifting instructions and status information), signal value (32 bits, storing actual data content), timestamp (64 bits, acquisition time accurate to milliseconds) and status flag (8 bits, indicating data validity and processing priority). Standardized encoding enables data from different sources and types to be processed uniformly in the same system.

[0025] The standardized information packets are CRC32 checked and encrypted. The CRC32 check can detect whether errors occur during data transmission. The calculation process is to perform polynomial division on each byte in the data stream to generate a 32-bit check code. The receiving end recalculates and compares the same algorithm to verify data integrity. The encryption process uses the AES-128 algorithm to ensure that the data is not unauthorizedly accessed or tampered with during the transmission of industrial Ethernet. The checked and encrypted data is transmitted to the intelligent processing unit via industrial Ethernet to form the system status data.

[0026] In order to achieve comprehensive monitoring of the shaft hoisting system, a digital twin model of the signal collection point is established based on the collected status data. The digital twin is a virtual copy of a physical entity or system in the digital world, which is synchronized with the entity through real-time data mapping. In this solution, a virtual model is created for each signal collection point (including the operating console and the sound and light signal device), and the model contains equipment attributes (model, location, function, etc.) and real-time status parameters. These parameters are updated once a second to ensure that the virtual model is synchronized with the actual equipment status, thus forming a complete virtual mapping structure of the shaft hoisting system.

[0027] The system compares and analyzes the virtual mapping structure with the historical collected data to identify abnormal collection points. The comparative analysis uses statistical methods to calculate the deviation between the current state and the historical data. First, a benchmark data set under normal working conditions is established, which contains the mean and standard deviation of each parameter. Then the current collected data is compared with the benchmark data, and the Z score is calculated (the difference between the current value and the mean divided by the standard deviation). When the Z score exceeds the preset threshold (usually set to ±3, indicating a deviation of 3 standard deviations), it is determined to be an abnormal point. In this way, the system can identify the collection points with abnormal data, mark these points, and generate lifting operation instruction data and equipment status data.

[0028] For example, a mine shaft hoisting system has TH12 (C) type operating console and KXH15 (B) type sound and light signal installed at the upper wellhead, middle level and lower wellhead respectively. When the operator at the upper wellhead presses the slow down button, the operating console generates the original signal. After the S7-1200 controller collects this signal, it removes interference through the Butterworth low-pass filter, and then encodes it into a standardized information packet, including the signal source identification upper wellhead operating console 001, signal type hoisting instruction, signal value slow down, acquisition timestamp and normal status flag. The data is CRC32-checked and encrypted and transmitted to the intelligent processing unit via industrial Ethernet. The system then updates the status parameters of the upper wellhead operating console in the digital twin model and compares and analyzes them with historical data. When it is found that the response time of a sound and light signal at the middle level is abnormal (Z score is 4.2, which is beyond the normal range), the system immediately marks the device as an abnormal point, and passes the processed slow down instruction and the device status data containing the abnormal mark to the subsequent processing module.

[0029] In a specific embodiment, the marking module 102 is used to: Performing time synchronization marking on the lifting operation instruction data to obtain time sequence instruction data; The timing instruction data is partitioned according to the three-layer network architecture, and communication nodes are established through programmable controllers to obtain a network architecture including the bottom field layer, the middle control layer and the upper management layer; Perform bus configuration on the communication connection in the network architecture, establish communication links through dual redundant design, and obtain the network topology structure; The heartbeat mechanism monitors the health status of each node in the network topology and obtains a distributed decision-making mechanism; The distributed decision-making mechanism is logically mapped and converted into controller configuration parameters through flameproof, intrinsic safety and explosion-proof design to obtain the controller working mode; A node dynamic access rule base is established based on the controller working mode, and the network topology adaptive adjustment is achieved through the self-organizing network algorithm to obtain an intrinsically safe distributed control topology structure.

[0030] Specifically, the marking module 102 performs time synchronization marking on the lifting operation instruction data, and the process adopts the high-precision time protocol (PTP). PTP realizes time synchronization through a master-slave architecture. The master clock device in the system periodically sends synchronization messages to the slave clock device, and the slave clock device adjusts the clock according to the received message and delay calculation. In the shaft lifting system, the KXT131 (A) -X machine room display device serves as the master clock source and sends a synchronization message every 500 milliseconds. Each operating console and annunciator serves as a slave clock device to receive the synchronization signal and adjust the local clock. The time synchronization error is controlled within 1 millisecond, ensuring that equipment in different locations has a unified time reference.

[0031] After the timing instruction data is generated, it is partitioned and processed according to the three-layer network architecture. The three-layer network architecture is a hierarchical network design method. In the shaft hoisting system, it is specifically divided into: the bottom field layer, including TH12 (C) operating console and KXH15 (B) sound and light signal device and other equipment that directly interact with the physical environment; the middle control layer, composed of S7-1200 and ST40 programmable controllers, is responsible for logic control and data processing; the upper management layer, mainly KXT131 (A) -X machine room display device and monitoring computer, is responsible for system monitoring and decision-making. Different levels use different communication protocols and processing priorities. The field layer equipment has the highest priority for data collection, the control layer has the second highest priority for data processing and execution of control logic, and the management layer has the lowest priority for monitoring and decision-making. Through this hierarchical processing method, a large number of concurrent instructions can be effectively handled to ensure that key control instructions are processed in a timely manner.

[0032] After the network architecture is established, the bus configuration is performed for the communication connection. In the shaft hoisting system, two industrial bus technologies, PROFIBUS-DP and PROFINET, are used. PROFIBUS-DP is a high-speed fieldbus applied to the field layer, with a transmission rate of up to 12Mbps, supporting a multi-master station structure, and suitable for connecting field devices such as operating consoles and annunciators; PROFINET is a high-performance bus based on industrial Ethernet, with a transmission rate of up to 100Mbps, supporting real-time communication, and is mainly used to connect control layer and management layer equipment. In order to improve system reliability, a dual redundant design is used to establish a communication link, that is, each key node has two independent communication paths. When the main path fails, the system automatically switches to the backup path, and the switching time does not exceed 100 milliseconds to ensure uninterrupted communication.

[0033] In order to monitor the running status of network nodes, the heartbeat mechanism is used to monitor the network topology. The heartbeat mechanism is a simple and effective method for monitoring node status. Its working principle is: each node periodically sends heartbeat messages (including node ID, timestamp and status information) to other nodes in the network. If a node does not receive the heartbeat message from another node within a predetermined time, it is considered that the node may have failed. In this system, the heartbeat period is set to 500 milliseconds. If the heartbeat message is not received for 3 consecutive times (i.e. 1.5 seconds), the node is considered offline. Based on the monitoring results of the heartbeat mechanism, the system applies the improved Raft distributed consensus algorithm to achieve state consistency between nodes. The Raft algorithm divides the nodes in the system into three roles: leader, follower and candidate. It ensures the consistency of the system state through election and log replication mechanisms, and can work normally even when some nodes fail.

[0034] Considering the particularity of the mine environment, it is necessary to logically map the distributed decision-making mechanism with the safety design requirements of the Coal Mine Safety Regulations. This process involves three key aspects: explosion-proof design, for potentially explosive environments, the design of the equipment housing can withstand internal explosions and prevent the flame from spreading to the outside; intrinsic safety design, ensuring that the spark energy generated by electrical equipment in normal or faulty conditions is insufficient to ignite the surrounding flammable gas; explosion-proof design, preventing the equipment from becoming an ignition source through positive pressure ventilation or potting. These safety design requirements are converted into controller configuration parameters, including: current limit parameters (controlled within the intrinsic safety current range, generally less than 100mA), voltage limit parameters (controlled within the intrinsic safety voltage range, usually less than 12V), communication timeout parameters (set to 3 times the heartbeat cycle, i.e. 1.5 seconds), etc. A node dynamic access rule base is established based on the controller working mode, and the network topology adaptive adjustment is achieved through the self-organizing network algorithm. The node dynamic access rule base defines the identity authentication, function identification and resource allocation rules for new nodes to join the network. The self-organizing network algorithm is based on the principle of edge computing, allowing nodes in the network to autonomously adjust their connection relationship and working mode according to the current network status and task requirements. When a new console or annunciator is detected, the system automatically completes identity authentication and function mapping, assigns network addresses and communication parameters, and establishes a connection with existing nodes. Similarly, when a node is offline, the system replans the communication path to ensure network connectivity.

[0035] For example, during the operation of a mine shaft hoisting system, the upper shaft console issues a slow-up command, which is first added with an accurate timestamp (for example, 2025-04-21 10:15:32.456). The system distributes and processes the timestamped commands according to a three-layer architecture: the TH12(C) console at the bottom field layer generates the operation signal; the S7-1200 controller at the middle control layer receives the signal and performs logical processing; the KXT131(A)-X display device at the upper management layer performs monitoring and display. The command is transmitted from the console to the controller via the PROFIBUS-DP bus, and then from the controller to the display device via the PROFINET bus. Each communication link has a backup path to deal with the failure of the main path. During this process, the heartbeat mechanism continuously monitors the status of each node, and the controller sends a heartbeat message every 500 milliseconds. When the system detects that a signaler at the intermediate level has not responded to the heartbeat message for three consecutive times, it immediately determines that it is offline, applies the Raft algorithm to reallocate control tasks, and adjusts the communication path through the self-organizing network algorithm, transferring the information that originally needed to pass through the signaler to other paths for transmission, ensuring that the instructions can be executed safely and reliably, and at the same time recording and reporting the fault information.

[0036] In a specific embodiment, the input module 103 is used to: The time series features of the equipment status data are extracted, and the change trend of continuous multi-frame data is calculated through a sliding window of 256 sampling points and a 75% overlap rate to obtain a 64-dimensional equipment status feature vector; The device state feature vector is input into a node-linked perceptron network with a three-layer structure, where the input layer contains 64 neurons to receive the feature vector, the hidden layer contains 128 neurons with ReLU activation function, and the output layer contains 5 neurons using Softmax function corresponding to 5 lifting states. The state probability distribution is obtained through forward propagation calculation; Based on the state probability distribution, the Markov decision process analysis is performed to obtain the improvement state confirmation result; Match the lifting status confirmation result with the lifting equipment knowledge graph constructed according to the triple structure, calculate the operation decision path through the shortest path algorithm, and obtain the lifting operation sequence; Applying the priority sorting rule to the promotion operation sequence, the control instruction set is obtained through a 3×5 dimensional priority matrix mapping, where the first dimension represents the instruction type, the second dimension represents the promotion state, and the third dimension is the priority value; The instruction encapsulation is completed by adding a 32-bit execution timing mark, a 64-bit parameter array and a 16-bit CRC check code to the control instruction set to obtain the shaft lifting control instruction.

[0037] Specifically, the input module 103 extracts time series features and uses a sliding window of 256 sampling points to segment the device status data. The overlapping rate between adjacent windows is maintained at 75%, which can capture short-term change features and maintain data continuity. In each window, time domain features, frequency domain features and statistical features are extracted and combined to form a 64-dimensional device status feature vector. This dimensionality reduction process compresses the original high-dimensional time series data into a feature vector containing key information.

[0038] After the feature vector is extracted, it is input into the node-linked perceptron network for state recognition. The node-linked perceptron network is a neural network structure designed specifically for mining hoisting systems. It consists of three layers: the input layer contains 64 neurons, corresponding to the 64 dimensions of the feature vector; the hidden layer contains 128 neurons, using the ReLU activation function (when the input is greater than 0, the output is equal to the input, otherwise the output is 0), which can effectively handle nonlinear relationships; the output layer contains 5 neurons, corresponding to the five lifting states of fast up, fast down, slow up, slow down and emergency stop, and the Softmax function is used to convert the output into a probability distribution. The forward propagation process of the network is that the data is transformed from the input layer through the weight matrix and the activation function, and then passed to the hidden layer and the output layer in turn. The output of each neuron is transformed by the weight matrix as the input of the next layer of neurons. The 5 neurons in the output layer represent the probability of the system being in various lifting states.

[0039] After obtaining the state probability distribution, the Markov decision process is used for further analysis. The Markov decision process is a random dynamic system model that is particularly suitable for describing the state transition process of the lifting system. In this model, the state space S (including 5 lifting states), the action space A (executable control actions), the state transition probability P (the probability of transitioning from one state to another) and the reward function R (evaluating the quality of states and actions) are defined. The model constructs a 5×5 state transition matrix to record the transition rules between different lifting states. For example, the transition probability from slow up to fast up is higher, while the transition probability from fast up directly to fast down is lower. The posterior probability is calculated by the Bayesian formula, and the lifting state confirmation result is obtained by combining the currently observed state probability distribution and the prior transition probability.

[0040] After the lifting status is confirmed, it is necessary to match it with the lifting equipment knowledge graph to determine the best operation sequence. The knowledge graph is a semantic network structure that stores knowledge in the form of (entity-relationship-entity) triples. In the shaft lifting system, entities include various types of equipment (operating consoles, signalers, hoists, etc.) and operating status, and relationships include control relationships, dependency relationships, etc. The knowledge graph contains about 15 operation nodes, and the Dijkstra shortest path algorithm is used to find the optimal path from the current state to the target state in the graph. The algorithm starts from the starting node, gradually expands to adjacent nodes, calculates the cumulative path weight, and selects the path with the smallest weight as the final decision path. The weight value reflects the complexity, risk and time cost of the operation. In this way, the system can find the safest and most efficient operation sequence.

[0041] After the operation sequence is generated, the priority sorting rules need to be applied for further processing. The priority matrix is ​​a 3×5-dimensional structure. The first dimension represents the instruction type (safety, emergency, and regular), the second dimension represents the lifting state (fast up, fast down, slow up, slow down, and emergency stop), and the third dimension is the priority value (range 1-10, the larger the value, the higher the priority). In the matrix, the priority value of safety instructions (such as emergency stop) is set to 8-10, the priority value of emergency instructions (such as fault handling) is set to 5-7, and the priority value of regular instructions (such as normal lifting) is set to 1-4. The system determines the priority of each instruction by looking up the table, and sorts them from high to low according to priority, forming a set of control instructions with a clear execution order.

[0042] Finally, the control instruction set is packaged in a standardized manner to form executable shaft hoisting control instructions. Three key pieces of information are added during the packaging process: a 32-bit execution timing mark (containing 16-bit absolute time and 16-bit relative time, used to specify the precise execution time point of the instruction), a 64-bit parameter array (containing specific execution parameters such as speed, displacement, acceleration, etc.) and a 16-bit CRC checksum (generated using polynomial division to verify the integrity of the instruction during transmission). Standardized packaging ensures that the instructions can be correctly identified and executed by the execution unit, while improving transmission reliability through CRC checking.

[0043] For example, during the operation of a mine shaft hoisting system, various sensors collect a set of equipment status data, including the status of the operating console buttons, hoist speed, position signals, etc. The input module first applies a 256-point sliding window to this set of data for time series feature extraction and calculates a 64-dimensional feature vector of the equipment operation status. These feature vectors are input into the node-linked perceptron network. The 64 neurons in the input layer of the network receive the feature data, which are then transferred to the 128 hidden layer neurons after weight matrix transformation. Each hidden layer neuron is processed by the ReLU activation function and then transferred to the 5 neurons in the output layer. Finally, the network outputs a set of probability values: [0.05, 0.03, 0.85, 0.06, 0.01], indicating that the probability that the system is currently in the slow-up state is 85%. This probability distribution is combined with the state transition matrix of the Markov decision process to confirm that the current hoisting state is slow-up. The system then queries the knowledge graph to determine that the shortest path from slow-up to the target state is: slow-up->deceleration->stop, which requires the execution of instructions of three operation nodes. These instructions are mapped through the priority matrix to obtain priority values, where the deceleration instruction belongs to the general class of instructions with a priority value of 3; the stop instruction also belongs to the general class of instructions with a priority value of 4. After the system sorts by priority, it adds an execution timing mark, a parameter array, and a CRC checksum to each instruction to form a complete control instruction package, which is finally sent to the execution unit for sequential execution.

[0044] In a specific embodiment, the implementation module 104 is used to: The shaft hoisting control instructions are parsed into execution action sequences, and a dependency model including three layers of execution units, namely, upper wellhead, middle level, and lower wellhead, is constructed through a directed acyclic graph structure to obtain the triggering conditions and execution order of each execution unit. Apply the horizontal forwarding interlocking rule to the dependency model, and through the mechanism of locking other signals at the same level until the current signal is confirmed, set the lower wellhead to only send signals to the upper wellhead, and the logical relationship that the current level interlocking is completed after the upper wellhead receives the signal sent by the lower wellhead, and obtain the horizontal forwarding interlocking structure that prevents signal conflicts; Based on the horizontal forwarding blocking structure, a hierarchical execution strategy is implemented. By dividing the execution process into three stages: preparation, execution, and completion, and setting a status check cycle for each stage, the execution control process is obtained; Add a double confirmation mechanism to the execution control process. By requiring two confirmation mechanisms, namely, instruction receipt feedback and execution completion feedback, and setting a timeout threshold for each confirmation, an instruction execution guarantee mechanism is obtained. Based on the instruction execution guarantee mechanism, the response data of each execution unit is collected, the device status, environmental parameters and operation results during the execution process are recorded, and real-time execution feedback data including timestamp, device ID, operation type and execution results is obtained; The real-time execution feedback data is stored in a three-level database architecture. A data backup mechanism is established through local cache, regional storage and central warehouse, and a rolling update strategy is set to obtain a record of the improvement process.

[0045] Specifically, the implementation module 104 parses the shaft hoisting control instruction generated by the input module 103 into a specific execution action sequence. The parsing process is carried out by the instruction decoder, which separates and extracts the execution timing mark, parameter array and check code in the instruction package, and verifies the check code to ensure the integrity of the instruction. After the instruction is parsed, the dependency model between the execution units is constructed through the directed acyclic graph structure. A directed acyclic graph is a directed graph structure that does not contain loops, which is very suitable for representing execution processes with sequential dependencies. In the shaft hoisting system, the nodes of the graph represent execution units at different positions, including the operating consoles and signalers at the upper wellhead, the middle level and the lower wellhead; and the edges of the graph represent the triggering relationship between the execution units, specifying the direction and conditions of signal transmission. The directed acyclic graph is analyzed by a topological sorting algorithm to determine the triggering conditions and execution order of each execution unit to form a complete execution plan.

[0046] After the dependency model is established, it is necessary to apply the horizontal forwarding interlocking rules for constraints. Horizontal forwarding interlocking is a safety mechanism in the mine hoisting system, which is used to prevent multiple operating points from sending conflicting signals at the same time. The specific implementation method is: when a horizontal operating station sends a signal, other operating stations on the same level will be temporarily locked until the signal processing is completed; at the same time, the signal transmission follows strict directional rules, and the lower wellhead can only send signals to the upper wellhead, but not to the same level or the lower level. This mechanism is implemented through a state lock table, which records the locking status and locking reason of each level. When the upper wellhead receives the signal sent from the lower wellhead, the system automatically checks and updates the locking status to ensure that the signal is transmitted along the predetermined path to prevent signal conflicts and confusion.

[0047] Based on the horizontal forwarding interlocking structure, a hierarchical execution strategy is implemented to finely control the execution process. The hierarchical execution strategy divides each execution action into three stages: preparation, execution, and completion. Each stage has clear start and end conditions. The preparation stage mainly performs resource checks and parameter verification to ensure that the execution environment meets the requirements; the execution stage is the actual execution process of the core operation; the completion stage is to confirm the results and release resources. A 5-millisecond status check cycle is set for each stage. The system regularly checks the execution status and records the progress. Only when the end conditions of a stage are met can it enter the next stage. This hierarchical execution method ensures that complex operations can be carried out in an orderly manner according to the predetermined steps.

[0048] To further improve the reliability of execution, a double confirmation mechanism is added to the execution control process. Double confirmation is a means to enhance the reliability of information transmission, requiring each execution instruction to be confirmed at two key points: instruction reception confirmation and execution completion confirmation. When the execution unit receives the instruction, it first sends a reception confirmation message to indicate that the instruction has been correctly received; after the execution is completed, it sends an execution completion confirmation message to indicate that the instruction has been successfully executed. To prevent the loss or delay of confirmation information, the system sets a timeout threshold of 25 milliseconds for each confirmation. If the confirmation information is not received within the timeout period, the retransmission mechanism or exception handling process will be triggered. This double confirmation mechanism greatly reduces the uncertainty in the instruction execution process and improves the reliability of the system.

[0049] Based on a complete instruction execution guarantee mechanism, the system can comprehensively collect the response data of each execution unit. Data collection is carried out at a frequency of 10 Hz, covering three types of key information: device status information, including device working mode, power status, communication status, etc.; environmental parameter information, including temperature, humidity, gas concentration and other safety-related parameters; operation result information, including operation type, execution status, completion degree, etc. The collected data is standardized to form a structured data packet containing four core fields: timestamp, device ID, operation type and execution result, as real-time execution feedback data.

[0050] To ensure the secure storage and efficient access of data, real-time execution feedback data is stored in a three-level database architecture. The three-level database architecture includes three levels: local cache, regional storage, and central warehouse. The local cache is located near each execution unit and uses memory database technology to store high-frequency access data in the last 30 minutes, providing a millisecond response speed; the regional storage is located in the control center of each level and uses a relational database to store regional data in the last 24 hours, providing a second-level query response; the central warehouse is located in the ground control center and uses distributed database technology to store complete historical data and support complex queries and data analysis. The system sets a 7-day rolling update strategy to automatically archive and compress data that exceeds the retention period, which not only ensures the integrity of the data but also optimizes the use of storage space.

[0051] For example, when a mine shaft hoisting system is performing a hoisting operation, the input module generates a slow-up control instruction. The implementation module first parses the instruction and extracts the execution timing mark to show that it should be executed at 10:15:30. The parameter array contains the speed setting of 1.5 m / s, the acceleration of 0.2 m / s², and the target position of 10 meters above the ground. The system then establishes an execution dependency graph and determines the execution order as follows: the lower wellhead operating console sends a signal → the middle level forwards the signal → the upper wellhead confirms the signal → the hoist executes the action. After applying the horizontal forwarding locking rule, when the lower wellhead sends a slow-up signal, the other operating consoles on the level are temporarily locked, and the signal is transmitted to the middle level and forwarded to the upper wellhead. After receiving the signal, the upper wellhead confirms and releases the locking state of the lower wellhead, and then sends a control command to the hoist. The entire execution process is divided into three stages: the preparation stage checks the hoist status and the safety door status; the execution stage controls the hoist to run upward at a speed of 1.5 m / s; the completion stage decelerates and stops after reaching the target position. Each execution unit sends a receipt confirmation immediately after receiving the instruction, and sends a completion confirmation again after the execution is completed. The system collects the response data of each execution unit at a frequency of 10 Hz throughout the process, records key parameters such as hoist speed, position, motor current, and stores these data in real time in a three-level database. Recent data is stored in the local cache, daily queries are performed through regional storage, and long-term data is archived to the central warehouse to form a complete record of the lifting process.

[0052] In a specific embodiment, the matching module 105 is used to: Extract mining features from real-time execution feedback data, and obtain the operation feature data of the mine hoisting system by calculating the feedback signal of the sound and light signal device, the status data of the operating console and the response time of the execution unit; Compare the mine hoisting system operation characteristic data with the preset mining fault characteristic library, and obtain the abnormal signal recognition result by calculating the matching degree with the common mining equipment failure mode; Determine the fault category based on the abnormal signal recognition results, and obtain the fault classification of mining hoisting equipment by identifying the sound and light signal fault, operating console fault, transmission line fault and controller fault; The fault classification of mining hoisting equipment is carried out to locate the fault area and equipment identification by analyzing the equipment status data of the upper wellhead, middle level and lower wellhead; The impact on the lifting operation is assessed based on the fault area and equipment identification. By judging whether it affects the transmission of lifting signals, blocks the horizontal forwarding function, or causes the emergency stop signal to be falsely triggered, the fault is divided into three levels: low, medium, and high, and the fault risk rating is obtained. A handling strategy is formulated based on the fault risk rating, including maintaining operation and planned repairs for low-level faults, repairing intermediate faults after the current upgrade cycle is completed, and immediate shutdown for high-level faults, resulting in a graded handling strategy.

[0053] Specifically, the matching module 105 performs mining feature extraction on the real-time execution feedback data. The feature extraction process focuses on the characteristics of the shaft hoisting system and analyzes three types of key data: the feedback signal of the sound and light signal, including parameters such as sound output status, light signal flashing frequency, brightness change, etc.; the state data of the operating console, including information such as button trigger status, display module working status, power supply voltage, etc.; the response time of the execution unit, recording the delay time from the issuance of the instruction to the response of the execution unit. The extraction method adopts the sliding time window technology, with 10 seconds as a window, sliding once every 5 seconds, and calculating the time series statistical features within the window, including the mean, standard deviation, maximum value, minimum value and change rate. For the sound and light signal, the periodic characteristics of the signal are additionally calculated to detect whether there is an abnormal signal interruption or frequency offset; for the operating console, the time correlation between the button trigger and the display response is analyzed; for the execution unit, the response time is compared with the preset normal response time range to identify abnormal delays. Through these feature extraction and calculation, an operation feature data set of the mine hoisting system is formed, which contains the working status characteristics and performance indicators of each component of the system.

[0054] After the feature data is extracted, it is compared with the preset mining fault feature library. The mining fault feature library is a reference database designed for mine hoisting systems, which contains feature descriptions of various common faults. The fault modes in the library are classified according to the equipment type and the nature of the fault, covering various types such as sound and light signal failure (such as sound failure, abnormal flashing of light signals), console failure (such as button stuck, display abnormality), transmission line failure (such as signal interruption, severe interference) and controller failure (such as program abnormality, memory overflow). The comparison process uses a similarity calculation method to calculate the Euclidean distance, cosine similarity and other indicators between the extracted operating feature data and the feature vectors of each fault mode in the library to generate a matching score. The Euclidean distance calculates the straight-line distance between feature vectors, and the cosine similarity focuses on the directional consistency of the feature vectors. The system sets a similarity threshold. When the matching degree exceeds the threshold, it is determined to be an abnormality, and the abnormal signal recognition result is output, including the possible fault type and matching score.

[0055] Based on the abnormal signal recognition results, the system further determines the specific fault category. The decision tree algorithm is used in the judgment process to construct a mapping relationship from abnormal features to specific fault types. Each node of the decision tree represents a feature judgment condition, and each path corresponds to a fault diagnosis logic. For example, when the light signal intensity of the sound and light signal device is detected to be abnormally low, but the sound output is normal and the power supply voltage is within the normal range, it is determined that the sound and light signal device light source is faulty. After layer-by-layer screening, the decision tree finally classifies the fault into four categories: sound and light signal fault, console fault, transmission line fault or controller fault. Each type of fault is also subdivided into multiple specific subcategories, such as sound and light signal fault is divided into sound module fault, light source fault, power supply fault, etc.; console fault is divided into button fault, display module fault, communication module fault, etc. Through this detailed classification, the system can accurately identify the specific type of fault and provide a clear direction for subsequent processing. After the fault category is determined, accurate location positioning is required to clarify the specific area and equipment where the fault occurs. The positioning process is achieved by analyzing the equipment status data from different areas. The system compares and analyzes the equipment status data of the upper wellhead, middle level and lower wellhead to find the source of the abnormal data. The analysis methods include signal propagation path tracking and abnormal feature correlation analysis. Signal propagation path tracking is to check the status data of each node along the signal transmission direction to find the node where the abnormality first occurs; abnormal feature correlation analysis is to find the correlation of abnormal features between devices in different areas. If devices in multiple areas show similar abnormalities, it may be caused by a common upstream fault. Through these analyses, the system can determine the location of the fault, accurate to the specific level and device, and output the positioning result including the coordinates of the fault area and the unique identifier of the device.

[0056] After the fault location is completed, the system evaluates the impact of the fault on the lifting operation based on the fault area and equipment identification. The evaluation process considers three key factors: whether it affects the lifting signal transmission, check whether the faulty equipment is located on the key signal transmission path, and whether the nature of the fault will cause signal loss or error; whether it blocks the horizontal forwarding function, determine whether the fault will interrupt the signal forwarding between different levels and affect the coordination of the entire lifting system; whether it causes the emergency stop signal to be falsely triggered, and judge whether the fault may cause the system to falsely trigger the emergency stop mechanism, causing unnecessary downtime. Based on the comprehensive evaluation of these three factors, the system divides the fault into three levels: low-level faults, which have a slight impact on the lifting operation and will not cause signal transmission problems or horizontal forwarding function disorders; intermediate faults, which will partially affect the lifting operation, such as causing some signal transmission delays or instability, but will not completely interrupt the operation; high-level faults, which seriously affect the lifting operation, such as blocking the key signal transmission path or causing the emergency stop signal to be falsely triggered. The evaluation results are output in the form of fault risk ratings to provide a decision-making basis for subsequent processing.

[0057] Finally, the system formulates corresponding processing strategies according to the fault risk rating. The processing strategy follows the principle of safety first and efficiency, and formulates differentiated processing solutions for faults of different risk levels: low-level faults adopt the strategy of maintaining operation and planning maintenance, the system continues to operate normally, and arranges equipment maintenance in the next maintenance cycle; intermediate faults adopt the strategy of repair after completing the current lifting cycle, allowing the system to complete the ongoing lifting operation, but requiring immediate maintenance after the current cycle ends; high-level faults adopt the strategy of immediate shutdown processing, the system triggers the safety shutdown procedure, interrupts the current lifting operation, and starts the emergency maintenance process. The processing strategy also includes specific execution details, such as maintenance personnel scheduling, spare parts preparation, alternative path configuration, etc., to ensure that the fault can be handled in a timely and effective manner. The final output of the hierarchical processing strategy contains information such as processing methods, execution time, resource requirements, and expected recovery time, providing clear guidance for system operation and maintenance.

[0058] For example, during the operation of a mine shaft hoisting system, the matching module collected abnormal feedback data of the KXH15 (B) type sound and light signal device at the middle level, which was manifested as unstable light signal flashing frequency and prolonged response time. The system first extracted the operating characteristic data of the sound and light signal device, including the change of light signal flashing frequency (from the normal 1 time / second to 0.7-1.3 times / second), response time (extended from the normal 50 milliseconds to 120 milliseconds) and power supply voltage (maintained within the normal range of 12V±0.5V). These characteristic data were compared with the mining fault feature library, and it was found that the matching degree with the fault mode of the sound and light signal device drive circuit reached 0.82, exceeding the threshold of 0.75, so it was determined that there was an abnormality. Through decision tree analysis, the system confirmed that this was a typical sound and light signal fault, and the specific subclass was a flash control circuit fault. Position positioning showed that the fault occurred on the No. 3 sound and light signal device at the middle level. By analyzing the position of the device in the horizontal forwarding chain, it was determined that it was responsible for forwarding the signal from the lower wellhead to the upper wellhead. Based on this positioning result, the system assessed that the fault would partially affect the transmission of the lifting signal, but because the system was designed with redundant transmission paths, it would not completely block the horizontal forwarding function and would not cause the emergency stop signal to be falsely triggered, so it was classified as a medium risk. According to the handling strategy for medium-level faults, the system decided to allow the current lifting operation to be completed, but after the end of this lifting cycle, the maintenance personnel were immediately arranged to replace the flash control circuit board of the sound and light signal device, and the backup signal transmission path was temporarily activated to ensure that the system can still operate normally during the maintenance period.

[0059] In a specific embodiment, the association module 106 is used to: Conduct statistical analysis on the fault handling effect in the hierarchical handling strategy, and obtain fault handling effect evaluation data by calculating the repair time, recovery success rate and recurrence frequency of different fault types; The fault handling effect evaluation data is time-series associated with the improvement process records, and a list of key parameters is obtained by matching the change trends of system operating parameters before and after the fault occurs. Conduct sensitivity analysis on the key parameter list, determine the influence weight of each parameter on system stability through single factor change test, and obtain the parameter influence ranking result; Based on the parameter influence ranking results, the control parameters are adjusted and calculated, and the optimal value range of each parameter is determined by setting the safety margin coefficient to obtain the optimized value of the control parameter; Logically analyze the locking events in the lifting process records, and obtain the optimization direction of the locking logic by extracting the association rules between the locking triggering conditions and the locking release conditions; Based on the optimization direction of the interlocking logic, a safety interlocking logic adjustment scheme is designed, and the signal transmission priorities, interlocking triggering conditions and interlocking release timing of the upper and lower wellheads and the intermediate levels are configured to obtain the safety interlocking logic adjustment scheme.

[0060] Specifically, the association module 106 performs a comprehensive statistical analysis on the fault handling effect in the hierarchical processing strategy generated by the matching module. The statistical analysis process revolves around three core indicators: repair time, which records the length of time from fault discovery to complete repair, and is subdivided into three sub-stages: fault confirmation time, maintenance preparation time, and actual repair time; recovery success rate, which calculates the proportion of normal operation of the equipment after the fault is repaired, and determines whether the repair is successful by analyzing the operating data within one week after the repair; recurrence frequency, which counts the number of times the same equipment or the same type of fault recurs within a specific time window. During the analysis, the fault type (sound and light signal fault, console fault, transmission line fault, and controller fault) and fault level (low, medium, and high) are classified, and the sliding time window technology is used, with one month as the window, sliding once a week, to calculate the time trend of various indicators. Through these statistical analyses, an evaluation data table containing various fault handling effects is generated, reflecting the processing efficiency and long-term reliability of different types of faults.

[0061] After the fault handling effect evaluation data is generated, it is necessary to conduct time series correlation analysis with the promotion process records to find the potential correlation between the fault and the system operation parameters. Time series correlation is a data mining technology used to discover the correlation and causal relationship between different time series data. The specific implementation method is to mark the fault event on the time axis, and then analyze the change trend of the system operation parameters in a certain time window before and after the fault occurs (usually 24 hours before the fault to 72 hours after the fault). The analysis process uses two methods: lag correlation analysis and Granger causality test. Lag correlation analysis calculates the correlation coefficient between the parameter and the fault under different time delays to find the time delay with the maximum correlation; Granger causality test determines whether the historical data of a parameter can significantly improve the prediction ability of the occurrence of faults by establishing an autoregressive model. Through these analyses, the system screens out the system operation parameters that are highly correlated with the fault and forms a list of key parameters, which usually include communication delay time, signal strength fluctuation, power supply voltage stability, ambient temperature and humidity, and other parameters. After the key parameters are determined, sensitivity analysis is performed on them to quantitatively evaluate the impact of each parameter on system stability. The sensitivity analysis adopts a single-factor change test method, that is, under the condition of keeping other parameters unchanged, the target parameter is slightly adjusted to observe the changes in the system stability index. The specific operation is to increase or decrease each key parameter within its normal value range in a 5% step in the test environment, and record the change value of the system stability index after each adjustment. System stability indicators include signal transmission success rate, instruction execution accuracy, system response time, etc. By calculating the change in stability index caused by each unit parameter change, the influence sensitivity of each parameter is obtained. The higher the sensitivity value, the greater the impact of the parameter on system stability. Finally, the parameters are sorted from high to low according to sensitivity, and the parameter influence ranking results are generated to provide a basis for subsequent parameter optimization.

[0062] Based on the ranking results of parameter influence, the system adjusts and calculates the control parameters to determine the optimal parameter value range. The core of the adjustment calculation is to maximize system performance while ensuring system safety. The concept of safety margin coefficient is introduced in the calculation process, that is, a certain margin is reserved to ensure safety on the basis of the theoretical optimal value. Different safety margin coefficients are set for parameters of different importance: the safety margin coefficient of key safety parameters is larger, usually 1.5-2.0; the safety margin coefficient of general performance parameters is smaller, usually 1.1-1.3. The specific adjustment method is: first determine the theoretical optimal value of each parameter based on the sensitivity analysis results, and then multiply it by the corresponding safety margin coefficient to obtain the recommended value in practical applications. For example, if the theoretical optimal timeout of a communication parameter is 20 milliseconds and the safety margin coefficient is 1.5, the actual set timeout should be 30 milliseconds. In this way, the system generates a set of control parameter optimization value tables containing the optimal value range of each control parameter to guide subsequent system configuration adjustments.

[0063] In addition to parameter optimization, the system also needs to perform special logic analysis on the locking events in the lifting process records to explore optimization opportunities in the locking mechanism. The locking event analysis first extracts all locking-related events from the lifting process records, including locking trigger events and locking release events. Then, these events are subjected to timing analysis and pattern mining to identify common trigger-release sequences and abnormal situations. Timing analysis evaluates the efficiency of the current locking logic by calculating indicators such as the duration of the locking state and the locking frequency; pattern mining uses association rule learning algorithms to discover the regular associations between locking trigger conditions and release conditions. Through these analyses, the system can identify the problem points in the current locking logic, such as unnecessary locking triggers, excessively long locking durations, and overly complex locking release conditions, thereby determining the optimization direction of the locking logic and providing clear goals for subsequent adjustments.

[0064] Based on the optimization direction of the interlocking logic, a detailed safety interlocking logic adjustment plan is designed. The adjustment plan focuses on three aspects: signal transmission priority configuration, assigning reasonable priorities to signals of different levels and types to ensure that key signals can be transmitted first; interlocking trigger condition setting, simplifying and refining interlocking trigger conditions, reducing the probability of false triggering and missed triggering; interlocking release timing optimization, clearly defining the conditions and processes for interlocking release, and shortening unnecessary interlocking duration. In specific implementation, first set the basic priority for various signals at the upper wellhead, middle level and lower wellhead (such as emergency signal 10, conventional lifting signal 7, status query signal 4), and then dynamically adjust according to the current system operation status (such as when the lifting operation is in progress, the priority of the relevant signal is increased by 2 points). The interlocking trigger condition adopts a three-stage design: precondition (the system is in a specific state), trigger event (specific signal issuance or specific operation execution) and confirmation condition (related equipment status change or feedback signal reception). The interlocking is triggered only when the three conditions are met at the same time. The lock release also uses a combination of multiple conditions, but adds an automatic timeout release mechanism. Even if certain conditions are not met, the lock is automatically released after the preset maximum lock time has expired, preventing the system from being in a locked state for a long time.

[0065] The above describes the shaft hoisting system based on intelligent control in the embodiment of the present application. The following describes the shaft hoisting system based on intelligent control in the embodiment of the present application. Figure 2 In the embodiments of the present application, an embodiment of a shaft hoisting system based on intelligent control includes: S201, collecting the operation console signal and the state information of the sound and light signal device at each operation point in the mine shaft hoisting system, and obtaining the hoisting operation instruction data and the equipment state data; S202, timestamping the lifting operation instruction data, building a field control network through a controller, and obtaining an intrinsically safe distributed control topology structure; S203, inputting the equipment status data into a lifting status analysis model based on a node-linked sensor network, identifying the lifting status, and obtaining a shaft lifting control instruction; S204, establishing a trigger dependency diagram between execution units according to the shaft hoisting control instruction, implementing a hoisting signal linkage control with a horizontal forwarding locking function, and obtaining real-time execution feedback data and hoisting process records; S205, extracting features and performing pattern matching on the real-time execution feedback data, identifying and locating abnormal states based on a preset fault feature library, and obtaining a hierarchical processing strategy; S206. Perform correlation analysis on the hierarchical processing strategy and the lifting process record to obtain the control parameter optimization value and the safety locking logic adjustment plan.

[0066] In the present application, by collecting signal data from each operating point of the mine shaft hoisting system, multi-level collaborative control is achieved, which significantly improves the safety, reliability and operating efficiency of the hoisting system; the hoisting operation instruction data is timestamped and an intrinsically safe distributed control topology is constructed, which solves the problems of unstable signal transmission and poor adaptability to the harsh underground environment in traditional systems, and ensures accurate timing coordination between execution units; the hoisting state analysis model based on the node-linked sensor network can accurately identify the five hoisting states of fast up, fast down, slow up, slow down and emergency stop, and converts uncertain sensor data into clear control instructions, reflecting the technical contribution of artificial intelligence algorithms in specific application fields, especially through the three-layer neural network structure to achieve high-precision state recognition, providing reliable guarantee for hoisting control in complex mine environments; the implementation of the dependency graph and the horizontal forwarding locking function effectively solves the signal conflict problem in the multi-level hoisting system, especially the situation that the lower wellhead can only send signals to the upper wellhead. The unidirectional control logic of the signal simplifies the system operation mechanism while improving safety. The feature extraction and fault matching functions of real-time execution feedback data realize early detection and precise positioning of faults, reducing the risk of equipment damage and downtime. The formulation of the graded processing strategy takes into account both safety and operational efficiency, and adopts differentiated processing measures according to the fault risk level. The control parameter optimization values ​​and safety interlock logic adjustment schemes obtained through correlation analysis enable the system to have self-optimization capabilities, and can continuously adjust the working parameters and control logic according to the operating data to adapt to the challenges brought by changes in the mine environment and aging of equipment. It is particularly worth emphasizing that the artificial intelligence algorithm in this scheme not only improves the accuracy of fault diagnosis through feature matching, but also clarifies the direction of improvement through sensitivity analysis in the parameter optimization process, reflecting the targeted contribution of the algorithm model in specific application scenarios, transforming traditional empirical judgments into data-driven decisions, and significantly improving the intelligence level and operation effect of the mine hoisting system.

[0067] above Figure 2 The shaft hoisting system based on intelligent control in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The shaft hoisting equipment based on intelligent control in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0068] Figure 3: is a structural schematic diagram of a shaft hoisting device based on intelligent control provided by an embodiment of the present invention. The shaft hoisting device 300 based on intelligent control may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 may be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the shaft hoisting device 300 based on intelligent control. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the shaft hoisting device 300 based on intelligent control to implement the steps of the shaft hoisting system based on intelligent control.

[0069] The shaft hoisting device 300 based on intelligent control may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The structure of the shaft hoisting equipment based on intelligent control shown does not constitute a limitation on the shaft hoisting equipment based on intelligent control provided by the present invention, and may include more or less components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0070] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the shaft hoisting system based on intelligent control.

[0071] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0072] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a shaft hoisting device based on intelligent control (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0073] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A shaft hoisting system based on intelligent control, characterized in that: The system includes: The acquisition module is used to collect the operation console signals and the status information of the sound and light signal devices at each operation point in the mine shaft hoisting system to obtain the hoisting operation instruction data and equipment status data; A marking module, used for timestamping the lifting operation instruction data, constructing a field control network through a controller, and obtaining an intrinsically safe distributed control topology structure; An input module, used for inputting the equipment status data into a lifting status analysis model based on a node-linked sensor network, identifying the lifting status, and obtaining a shaft lifting control instruction; An implementation module, used to establish a trigger dependency diagram between execution units according to the shaft hoisting control instruction, implement the hoisting signal linkage control with horizontal forwarding locking function, and obtain real-time execution feedback data and hoisting process records; A matching module, used to perform feature extraction and pattern matching on the real-time execution feedback data, identify and locate abnormal states based on a preset fault feature library, and obtain a hierarchical processing strategy; The correlation module is used to correlate and analyze the hierarchical processing strategy and the lifting process record to obtain the control parameter optimization value and the safety locking logic adjustment plan.

2. The shaft hoisting system based on intelligent control according to claim 1 is characterized in that: Acquisition module, used for: An operating table and an acoustic and optical signal device are installed at the upper wellhead, middle level and lower wellhead of the shaft respectively. Signals are collected through the controller to obtain the original data including lifting instructions and equipment status. The raw data is subjected to noise filtering through a Butterworth low-pass filter to obtain a boost operation signal; Digitally encoding the lifting operation signal according to the signal source identifier, signal type identifier, signal value, time stamp and status flag to obtain a standardized information packet; Performing CRC32 check and encryption processing on the standardized information packet, transmitting it to the intelligent processing unit via industrial Ethernet, and obtaining the lifting system status data; A digital twin model of the signal collection point is established according to the lifting system status data, and a virtual mapping structure of the shaft lifting system is obtained by updating the status parameters in real time; The virtual mapping structure is compared and analyzed with the historical collection data, and abnormal collection points are identified through deviation calculation to obtain enhanced operation instruction data and equipment status data.

3. The shaft hoisting system based on intelligent control according to claim 1 is characterized in that: Marking module for: Performing time synchronization marking on the lifting operation instruction data to obtain timing instruction data; The timing instruction data is partitioned according to a three-layer network architecture, and communication nodes are established through a programmable controller to obtain a network architecture including a bottom field layer, a middle control layer, and an upper management layer; Performing bus configuration on the communication connection in the network architecture, establishing a communication link through a dual redundant design, and obtaining a network topology structure; The health status of each node in the network topology is monitored by a heartbeat mechanism to obtain a distributed decision-making mechanism; The distributed decision-making mechanism is logically mapped and converted into controller configuration parameters through flameproof, intrinsic safety and explosion-proof design to obtain the controller working mode; A node dynamic access rule base is established based on the controller working mode, and the network topology is adaptively adjusted through a self-organizing network algorithm to obtain an intrinsically safe distributed control topology structure.

4. The shaft hoisting system based on intelligent control according to claim 1, characterized in that: Input modules for: Extracting time series features from the device status data, calculating the change trend of multiple frames of data through a sliding window of 256 sampling points and a 75% overlap rate, and obtaining a 64-dimensional device status feature vector; Input the device state feature vector into a node-linked perceptron network with a three-layer structure, wherein the input layer includes 64 neurons receiving the feature vector, the hidden layer includes 128 neurons with a ReLU activation function, and the output layer includes 5 neurons using a Softmax function corresponding to 5 boosted states, and the state probability distribution is obtained by forward propagation calculation; Performing a Markov decision process analysis based on the state probability distribution to obtain a boost state confirmation result; Matching the lifting state confirmation result with the lifting equipment knowledge graph constructed according to the triple structure, calculating the operation decision path through the shortest path algorithm, and obtaining the lifting operation sequence; Applying a priority sorting rule to the lifting operation sequence, mapping through a 3×5 dimensional priority matrix, where the first dimension represents the instruction type, the second dimension represents the lifting state, and the third dimension is the priority value, to obtain a control instruction set; The instruction encapsulation is completed by adding a 32-bit execution timing mark, a 64-bit parameter array and a 16-bit CRC check code to the control instruction set to obtain the shaft lifting control instruction.

5. The shaft hoisting system based on intelligent control according to claim 1, characterized in that: Implement modules for: The shaft hoisting control instruction is parsed into an execution action sequence, and a dependency model including three layers of execution units, namely, upper wellhead, middle level, and lower wellhead, is constructed through a directed acyclic graph structure to obtain the triggering condition and execution order of each execution unit; Apply the horizontal forwarding interlocking rule to the dependency model, and through the mechanism of locking other signals at the same level until the current signal is confirmed, set the lower wellhead to only send signals to the upper wellhead, and the logical relationship that the current level interlocking is completed after the upper wellhead receives the signal sent by the lower wellhead, and obtain the horizontal forwarding interlocking structure to prevent signal conflicts; Implementing a hierarchical execution strategy based on the horizontal forwarding blocking structure, by dividing the execution process into three stages: preparation, execution and completion, and setting a status check cycle for each stage, thereby obtaining an execution control process; A double confirmation mechanism is added to the execution control process, by requiring two confirmation mechanisms, namely, instruction reception feedback and execution completion feedback, and setting a timeout threshold for each confirmation, to obtain an instruction execution guarantee mechanism; Based on the instruction execution guarantee mechanism, the response data of each execution unit is collected, the device status, environmental parameters and operation results during the execution process are recorded, and real-time execution feedback data including timestamp, device ID, operation type and execution result is obtained; The real-time execution feedback data is stored in a three-level database architecture, a data backup mechanism is established through local cache, regional storage and central warehouse, and a rolling update strategy is set to obtain the improvement process record.

6. The shaft hoisting system based on intelligent control according to claim 1, characterized in that: Matching modules for: Extracting mining characteristics from the real-time execution feedback data, and obtaining the mine hoisting system operation characteristic data by calculating the feedback signal of the sound and light signal device, the operating console status data and the execution unit response time; The operation characteristic data of the mine hoisting system are compared with a preset mining fault characteristic library, and the abnormal signal recognition result is obtained by calculating the matching degree with the common mining equipment failure mode; Determine the fault category based on the abnormal signal recognition result, and obtain the fault classification of the mining hoisting equipment by identifying the sound and light signal fault, the operating console fault, the transmission line fault and the controller fault; Positioning the fault classification of the mining hoisting equipment, and obtaining the fault area and equipment identification by analyzing the equipment status data of the upper wellhead, the middle level and the lower wellhead; Based on the fault area and equipment identification, the impact on the lifting operation is evaluated. By judging whether it affects the transmission of lifting signals, whether it blocks the horizontal forwarding function, and whether it causes the emergency stop signal to be falsely triggered, the fault is divided into three levels: low, medium and high, and the fault risk rating is obtained; A handling strategy is formulated based on the fault risk rating, including maintaining operation and planned repairs for low-level faults, repairing intermediate faults after the current upgrade cycle is completed, and immediately shutting down for high-level faults, thereby obtaining a graded handling strategy.

7. The shaft hoisting system based on intelligent control according to claim 1, characterized in that: Association modules for: Performing statistical analysis on the fault handling effect in the hierarchical handling strategy, and obtaining fault handling effect evaluation data by calculating the repair time, recovery success rate and recurrence frequency of different fault types; Associating the fault handling effect evaluation data with the improvement process record in time series, and obtaining a key parameter list by matching the change trend of system operation parameters before and after the fault occurs; Perform sensitivity analysis on the key parameter list, determine the influence weight of each parameter on system stability through single factor change test, and obtain parameter influence ranking results; Based on the parameter influence ranking result, the control parameters are adjusted and calculated, and the optimal value range of each parameter is determined by setting the safety margin coefficient to obtain the optimized value of the control parameter; Performing logic analysis on the locking events in the lifting process record, and obtaining the optimization direction of the locking logic by extracting the association rules between the locking triggering conditions and the locking release conditions; Based on the optimization direction of the locking logic, a safety locking logic adjustment scheme is designed, and the signal transmission priorities, locking trigger conditions and locking release timing of the upper and lower wellheads and the middle level are configured to obtain the safety locking logic adjustment scheme.

8. A shaft hoisting method based on intelligent control, characterized in that: The shaft hoisting method based on intelligent control comprises: Collect the operation console signals and sound and light signal status information of each operation point in the mine shaft hoisting system to obtain the hoisting operation instruction data and equipment status data; The lifting operation instruction data is timestamped, and a field control network is constructed through a controller to obtain an intrinsically safe distributed control topology structure; Inputting the equipment status data into a lifting status analysis model based on a node-linked sensor network to identify the lifting status and obtain a shaft lifting control instruction; According to the shaft hoisting control instruction, a trigger dependency diagram between execution units is established, and a hoisting signal linkage control with a horizontal forwarding locking function is implemented to obtain real-time execution feedback data and hoisting process records; Perform feature extraction and pattern matching on the real-time execution feedback data, identify and locate abnormal states based on a preset fault feature library, and obtain a hierarchical processing strategy; The hierarchical processing strategy and the lifting process record are correlated and analyzed to obtain the optimized value of the control parameter and the safety locking logic adjustment plan.

Images