Coal cutter traction speed cross-cycle control method and system based on temperature field sensing

Abstract

The invention discloses a coal cutter traction speed cross-cycle control method and system based on temperature field sensing, and the method comprises the steps: collecting multi-dimensional information to construct a multi-source sensing data set during the Nth cutting; extracting thermal response characteristics of the temperature field, and obtaining coal rock properties at the current position through the identification model; constructing position-coal rock character space mapping and storing the position-coal rock character space mapping into a knowledge base as priori; the prior information of the last knife is called at the N + 1th knife, and mutual verification and correction are carried out in combination with real-time temperature field sensing; a multi-objective optimization model is constructed, and the optimal traction speed is decided under the condition that constraints are met; and the optimal speed instruction is issued and executed, and feedback parameters are continuously monitored, so that operation safety is ensured. According to the system, a sensing acquisition module, a thermal response feature extraction module, a coal rock character recognition module, a space mapping and knowledge base storage module, a prior calling and correction module, a multi-target traction speed optimization module and a traction control execution module are connected in sequence. According to the method and the system, the stability and the intelligent decision-making level of cutting operation can be improved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent control technology for fully mechanized mining, specifically relating to a method and system for cross-cycle control of the traction speed of a coal mining machine based on temperature field sensing. Background Technology

[0002] In fully mechanized coal mining operations, the traction speed of the coal mining machine plays a decisive role in cutting load, energy consumption level, and equipment operating safety. The coal and rock properties are unevenly distributed in the working face and continuously change as the coal mining machine advances, making the rational adjustment of the coal mining machine's traction speed a core challenge in the field of fully mechanized mining face control.

[0003] Currently, under the existing technological system, there are two main ways to set the traction speed of coal mining machines: one is to set it based on the operator's experience; the other is to adjust it based on feedback from single operating parameters such as motor current and power. These methods have significant drawbacks, as they cannot accurately reflect the dynamic changes in coal and rock properties. When facing complex coal and rock conditions, this can easily lead to increased fluctuations in traction load and a significant increase in energy consumption, which in turn can seriously affect the stability of the cutting process and the safety of equipment operation.

[0004] With the continuous development of intelligent sensing technology, some technical solutions attempt to identify coal and rock properties using coal face images or temperature field information, and adjust the traction speed accordingly. However, most of these methods rely on computationally intensive deep learning models. In actual underground working environments, there are many practical problems such as limited computing power, fluctuating communication bandwidth, and unstable latency. These problems make these methods place excessive demands on the onboard computing power and real-time communication of the coal mining machine, making them difficult to adapt effectively.

[0005] In addition, existing intelligent control schemes typically tightly couple coal and rock identification with traction control in the same real-time closed-loop system. This architecture, which is highly dependent on real-time computing power, reduces the system's engineering applicability and operational reliability under complex working conditions.

[0006] In summary, there is an urgent need for a control method that can realize coal and rock property perception and adaptive adjustment of coal mining machine traction under limited computing power and complex communication conditions, so as to effectively improve the system's adaptability to complex coal and rock conditions and meet the actual needs of safe underground coal mining operations. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention provides a cross-cycle control method and system for the traction speed of a coal mining machine based on temperature field sensing. This method is simple to implement and has low implementation cost. It can achieve adaptive rolling optimization of the traction speed of the coal mining machine, significantly improving the cutting stability and intelligent decision-making level under complex geological conditions while taking into account cutting efficiency and energy consumption. Through the collaborative work of multiple modules, the system realizes cross-cycle decoupling and rolling optimization of coal and rock property sensing and coal mining machine traction speed control, effectively improving the adaptability and operational stability of the coal mining machine in environments with limited computing power and complex communication conditions.

[0008] To achieve the above objectives, the present invention provides a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, comprising the following steps: S1: Data acquisition and status perception; during the Nth cut, coal wall temperature field information is acquired; coal mining machine operating status parameters are acquired simultaneously, and the position information of the coal mining machine in the working face is obtained to construct a multi-source perception dataset for the cutting process; S2: Thermal response feature extraction and coal and rock property identification; Based on the collected coal wall temperature field information, feature parameters reflecting the thermal response characteristics of coal and rock are extracted, and the coal and rock property identification and classification are completed through the preset coal and rock property identification model to obtain the coal and rock property identification results at the current location; S3: Spatial mapping construction and knowledge storage; The coal and rock property identification results are spatiotemporally correlated with the coal mining machine location information to construct a spatial mapping relationship between location and coal and rock properties, which is used to characterize the distribution characteristics of coal and rock properties along the working face in the current cutting cycle; At the same time, the spatial mapping information of coal and rock properties is stored in the knowledge base as a priori basis for subsequent cutting cycles. S4: Prior knowledge retrieval and real-time correction; During the N+1th cut, based on the real-time position of the coal mining machine, the corresponding coal and rock property spatial mapping information formed by the previous cut cycle is retrieved from the knowledge base to obtain the corresponding coal and rock property at the current position as prior input; At the same time, coal wall temperature field information is continuously collected to identify and monitor the actual changes in coal and rock properties in real time, realizing mutual verification and correction between prior knowledge and real-time perception, and outputting the corrected coal and rock property identification result; S5: Multi-objective constraint index speed decision; After obtaining the corrected coal and rock property identification results, construct a multi-objective constraint optimization model; Under the premise of satisfying the multi-objective constraint conditions, with the goal of improving cutting stability, reducing traction load fluctuations, reducing cutting energy consumption, and taking cutting efficiency into account, determine the optimal traction speed setting value that matches the current coal and rock properties. S6: Traction Execution and Process Control; The optimal traction speed setpoint is used as the control input, the optimal traction speed command is output and sent to the traction drive system, and the coal mining machine is driven to run smoothly according to the optimal traction speed setpoint; During the cutting process, real-time feedback parameters are continuously monitored to ensure that the operating status is always within the safe constraint range; S7: Knowledge Fusion Update and Cross-Cycle Iteration; After the N+1th cut is completed, the coal and rock property identification results obtained through real-time correction in the current cutting cycle are fused and updated with the original coal and rock property spatial mapping information in the knowledge base; The updated coal and rock property spatial mapping information is rewritten into the knowledge base as the prior basis for the next cutting cycle; By repeatedly executing S1 to S7 in multiple cutting cycles, the cross-cycle rolling optimization of the coal mining machine traction speed and the adaptive evolution of the system are realized.

[0009] As a preferred embodiment, in S1, the motion state parameters include traction speed, motor current, and power.

[0010] As a preferred option, in S1, the temperature field information of the coal wall is collected by an infrared thermal imaging device installed at the front of the coal mining machine.

[0011] As a preferred embodiment, in S2, the characteristic parameters include temperature gradient characteristic parameters and heat distribution characteristic parameters.

[0012] As a preferred option, in S2, the coal and rock properties include hardness, joints, and coal-rock interface information.

[0013] As a preferred option, in S6, the feedback parameters include traction load and motor current parameters.

[0014] As a preferred option, in S5, the multi-objective constraint conditions of the multi-objective constraint optimization model include at least the traction system load constraint, motor current / power constraint, and cutting safety constraint.

[0015] This invention provides a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing. First, it simultaneously collects the coal face temperature field, the operating parameters of the coal mining machine, and location information to construct a multi-source sensing dataset. Compared to relying on a single sensor, temperature field information can detect changes in coal and rock in advance (thermal response precedes changes in mechanical load), providing richer and earlier input for subsequent identification and decision-making. Simultaneously, location information and temperature field are collected synchronously, laying a precise spatiotemporal benchmark for constructing a spatial mapping relationship between location and coal and rock properties. Second, it utilizes the thermal response characteristics of coal and rock friction during cutting to achieve non-contact, real-time identification of coal and rock properties, avoiding the wear and failure problems of traditional contact sensors. Since thermal response characteristic parameters have an inherent physical correlation with the physical properties of coal and rock (hardness, brittleness, coefficient of friction), the extraction of thermal response characteristic parameters makes the identification results more interpretable and reliable, rather than a black-box model output. Next, the coal and rock property identification results are spatiotemporally correlated with the location of the coal mining machine to construct a spatial mapping relationship between location and coal and rock properties. This achieves a digital, visual, and continuous representation of coal and rock distribution characteristics, providing map-like prior knowledge for subsequent cutting and a stable and reliable prior basis for subsequent traction control. Simultaneously, for the first time, coal and rock property information is stored in a structured form in a knowledge base, giving the system memory capabilities and breaking the traditional independent, memoryless control mode of each cut of the coal mining machine. Subsequently, during the N+1 cut, the prior mapping (coal and rock map) from the previous cut is invoked as a decision guide, while real-time temperature field sensing is used to monitor and verify actual coal and rock changes. This mechanism of prior knowledge combined with real-time dual-channel verification effectively overcomes the problem of inaccuracy caused by sudden changes in coal and rock (such as faults or scour zones) due to a single prior knowledge, and also avoids misjudgments caused by local thermal interference due to a single real-time sensing. Through mutual verification between prior knowledge and real-time data, the corrected coal and rock property identification results are output, achieving continuous improvement in sensing accuracy as the cutting process progresses. Furthermore, by constructing a multi-objective constrained optimization model, multiple mutually restrictive objectives such as cutting stability, load fluctuation suppression, energy consumption reduction, and cutting efficiency are simultaneously considered. Compared to traditional single-objective control methods (such as constant power or constant speed), this method can achieve a more balanced and intelligent traction speed decision-making process under complex coal and rock conditions. It can effectively suppress traction load fluctuations, reduce mechanical specific energy, and significantly improve the stability and energy efficiency of the cutting process. The optimal traction speed setpoint is calculated in real time based on the current (corrected) coal and rock properties, which helps to achieve refined control that adapts to different rock conditions and adjusts speed accordingly.Then, after the optimal speed command is sent to the traction system, real-time feedback parameters (such as motor current, vibration, and temperature rise) are continuously monitored to ensure that the operating status remains within safe constraints. This approach embodies the composite control concept of feedforward and feedback, effectively responding to changes in coal and rock while ensuring equipment safety. Simultaneously, this method fully considers mechanical inertia and response time, avoiding mechanical shock conditions caused by sudden speed changes. Finally, after each cut, the updated data obtained from real-time correction is integrated with the original information in the knowledge base and rewritten into the knowledge base. This ensures that the knowledge base can continuously update and optimize itself as the working face advances, giving the system adaptive evolution capabilities. By repeatedly executing S1 to S7 in multiple cutting cycles, cross-cycle rolling optimization of traction speed decisions is achieved. Each cut is both a response to the current coal and rock conditions and a learning process for the next cut. This mechanism allows the system to gradually adapt to long-term coal seam changes (such as gradual changes in coal quality along the working face) and continuously evolve adaptively with the cutting process, improving the stability and reliability of the coal mining machine under complex coal and rock conditions. Meanwhile, by employing a cross-cycle priori guidance mechanism, the reliance on fully real-time coal and rock identification and high-performance computing capabilities is reduced, improving the system's engineering applicability in complex underground environments. This invention focuses on the reciprocating cutting operation of a coal mining machine in a fully mechanized mining face. Using the single-cut cutting process as the basic control cycle unit, a cross-cycle closed-loop control mechanism is constructed, encompassing perception, identification, mapping, priori guidance and real-time correction, traction optimization, execution, and rolling updates. During the Nth cut, the system emphasizes the perception and spatial mapping of coal and rock properties. During the N+1th cut, the system makes traction control decisions based on the coal and rock property distribution information from the previous cut and continuously corrects for changes in coal and rock through continuous perception. This iterative optimization of the control effect across multiple cutting cycles enables the coal mining machine's traction control process to continuously learn and optimize, overcoming the limitations of traditional fixed parameters or manual intervention.

[0016] This method is simple to implement and has low implementation costs. It can achieve adaptive rolling optimization of the traction speed of the coal mining machine, and while taking into account cutting efficiency and energy consumption, it significantly improves cutting stability and intelligent decision-making level under complex geological conditions.

[0017] This invention also provides a cross-cycle control system for the traction speed of a coal mining machine based on temperature field sensing, used to implement a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, comprising: The sensing and acquisition module is installed on the coal mining machine and is used to collect coal wall temperature field information, coal mining machine operating status parameters, and coal mining machine position information in the working face in a non-contact manner. A thermal response feature extraction module, which is connected to the sensing and acquisition module, is used to extract features from the coal wall temperature field information to obtain thermal response feature parameters; A coal and rock property identification module, which is connected to a thermal response feature extraction module, is used to identify and classify coal and rock properties based on thermal response feature parameters and output the coal and rock property identification results at the current location. A spatial mapping and knowledge base storage module is connected to a coal and rock property identification module. It is used to spatiotemporally associate the coal and rock property identification results with the coal mining machine location information, construct a location-coal and rock property spatial mapping relationship, and store the coal and rock property spatial mapping information in a knowledge base. The prior knowledge retrieval and correction module is connected to both the knowledge base and the sensing and acquisition module. In subsequent cutting cycles, it retrieves the corresponding coal and rock property spatial mapping information generated in the previous cutting cycle from the knowledge base based on the real-time position of the coal mining machine, using it as the prior input for the coal and rock properties at the current position. Simultaneously, it is used to identify and monitor changes in coal and rock properties in real time based on the real-time acquired coal wall temperature field information, achieving mutual verification and correction between prior knowledge and real-time sensing, and outputting the corrected coal and rock property identification result. A multi-objective traction speed optimization module is connected to both the sensing and acquisition module and the prior call and correction module. Based on the corrected coal and rock property identification results, the module determines the optimal traction speed setpoint through a multi-objective optimization algorithm under multi-objective constraints. Simultaneously, it monitors the coal mining machine's operating feedback parameters in real time and ensures that the cutting process is always within the safe constraint range, forming a control closed loop. The traction control execution module is used to take the optimal traction speed setpoint as the control input, make a decision and output the optimal traction speed command to the traction drive system, so as to realize the adaptive adjustment of the traction speed of the coal mining machine.

[0018] As a preferred embodiment, a rolling update module is also included. The rolling update module is used to merge and update the coal and rock property identification results obtained by real-time correction in the current cutting cycle with the original coal and rock property spatial mapping information in the knowledge base, generate updated coal and rock property spatial mapping information and rewrite it into the knowledge base as a priori basis for the next cutting cycle, thereby realizing cross-cycle rolling update of coal and rock property information.

[0019] In this technical solution, by setting up a rolling update module, not only is the control effect of a single cutting cycle improved, but more importantly, the system is given the vitality of continuous evolution, so that the traction control of the coal mining machine can continuously improve itself as the working face advances, and finally achieves true adaptive and intelligent operation.

[0020] As a preferred embodiment, the sensing and acquisition module includes an infrared thermal imaging device, a multi-parameter monitoring integrated device, and a positioning monitoring device; the infrared thermal imaging device is used to acquire coal face temperature field information; the multi-parameter monitoring integrated device is used to acquire coal mining machine operating status parameters; and the positioning monitoring device is used to acquire the coal mining machine's position information in the working face.

[0021] In this invention, the sensing and acquisition module enables non-contact sensing of coal wall temperature, avoiding sensor wear issues caused by direct contact with coal and rock, thus improving system reliability and lifespan. Simultaneously, the synchronous acquisition of multi-dimensional data provides a rich data foundation for subsequent accurate identification and control. The thermal response feature extraction module transforms raw, complex temperature field data into thermal response feature parameters with clear physical meaning, such as temperature change rate and thermal diffusivity. This process achieves a crucial transformation from data to features, significantly reducing the complexity of subsequent identification. Furthermore, this feature extraction process effectively filters out environmental noise and interference, highlighting temperature response changes caused by differences in the physical properties of coal and rock, improving information effectiveness and robustness of identification. The coal and rock property identification module leverages the inherent differences in thermal conductivity, thermal diffusivity, and other thermophysical properties between coal and rock for convenient and accurate identification. Compared to indirect methods relying on consequential parameters such as vibration and current during cutting, temperature field-based identification is more predictive and accurate. Therefore, the coal and rock properties at the current location can be classified and identified in real time during the movement of the coal mining machine, providing an immediate basis for dynamic adjustments. Through the setting of spatial mapping and knowledge base storage modules, the identified coal and rock properties are innovatively correlated with the precise location of the coal mining machine in time and space, constructing a spatial mapping relationship between location and coal and rock properties. This transforms the information perceived during the cutting process into structured knowledge for storage, making each cutting operation a process of accumulating experience for the next cutting operation, achieving knowledge accumulation and reuse. Through the setting of prior knowledge retrieval and correction modules, the historical experience in the knowledge base (prior information from the previous cycle) is creatively combined with the real-time perceived information of the current cycle. This dual-source information verification mechanism greatly improves the reliability and accuracy of coal and rock property identification. Simultaneously, when there is a deviation between real-time perception and prior knowledge, the system can correct based on real-time perception. This can not only cope with sudden changes in local geological conditions but also reverse-verify and optimize the knowledge base, enabling the knowledge base to have self-learning and dynamic update capabilities. By configuring a multi-objective traction speed optimization module, the results of coal and rock property identification and correction can be used as input. Within a unified framework, multiple, even conflicting, objectives such as production efficiency (high traction speed), equipment safety (low load), and energy consumption are comprehensively considered. An optimization algorithm is used to find the optimal, rather than the fastest, traction speed. Based on this, the operation feedback of the coal mining machine is monitored in real time to ensure that any optimization decision will not cause the equipment to operate beyond its safety boundaries. This constitutes a complete control closed loop of perception-decision-execution-feedback, guaranteeing the safe and stable operation of the system. Through the configuration of the traction control execution module, the optimal speed setpoint obtained from the optimization algorithm can be accurately and quickly converted into control commands that the drive system can recognize, ensuring that the upper-level intelligent decisions are accurately implemented in the lower-level physical world.This invention takes temperature field sensing as an innovative entry point, and combines spatial knowledge graphs and multi-objective optimization technology to transform passive interception into active perception and pre-adaptation, single control into multi-objective collaborative optimization, and single interception into cross-cycle self-learning, thus constructing an intelligent control ecosystem that integrates prediction, memory, optimization, and correction.

[0022] Through the collaborative work of multiple modules, the system achieves cross-cycle decoupling and rolling optimization of coal and rock property perception and coal mining machine traction speed control, effectively improving the adaptability and operational stability of the coal mining machine in environments with limited computing power and complex communication conditions. Attached Figure Description

[0023] Figure 1 This is a flowchart of the monitoring method in this invention; Figure 2 This is a block diagram illustrating the principle of the monitoring system in this invention. Detailed Implementation

[0024] The invention will now be further described with reference to the accompanying drawings.

[0025] like Figure 1 As shown, this invention provides a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, comprising the following steps: S1: Data acquisition and status perception; during the Nth cut, coal wall temperature field information is acquired; coal mining machine operating status parameters are acquired simultaneously, and the position information of the coal mining machine in the working face is obtained to construct a multi-source perception dataset for the cutting process; S2: Thermal response feature extraction and coal and rock property identification; Based on the collected coal wall temperature field information, feature parameters reflecting the thermal response characteristics of coal and rock are extracted, and the coal and rock property identification and classification are completed through the preset coal and rock property identification model to obtain the coal and rock property identification results at the current location, which can be used to reflect the physical properties of coal and rock. S3: Spatial mapping construction and knowledge storage; The coal and rock property identification results are spatiotemporally correlated with the coal mining machine location information to construct a spatial mapping relationship between location and coal and rock properties, which is used to characterize the distribution characteristics of coal and rock properties along the working face in the current cutting cycle; At the same time, the spatial mapping information of coal and rock properties is stored in the knowledge base as a priori basis for subsequent cutting cycles. S4: Prior knowledge retrieval and real-time correction; During the N+1th cut, based on the real-time position of the coal mining machine, the corresponding coal and rock property spatial mapping information formed by the previous cut cycle is retrieved from the knowledge base to obtain the corresponding coal and rock property at the current position as prior input to guide traction control decisions; At the same time, coal wall temperature field information is continuously collected to identify and monitor the actual changes in coal and rock properties in real time, realizing mutual verification and correction between prior knowledge and real-time perception, and outputting the corrected coal and rock property identification results to provide updated data for subsequent cuts; S5: Multi-objective constraint index speed decision; After obtaining the corrected coal and rock property identification results (prior + real-time correction), a multi-objective constraint optimization model is constructed; Under the premise of satisfying the multi-objective constraint conditions, with the goals of improving cutting stability, reducing traction load fluctuations, reducing cutting energy consumption, and taking cutting efficiency into account, the optimal traction speed setting value matching the current coal and rock properties is determined. S6: Traction Execution and Process Control; The optimal traction speed setpoint is used as the control input, the optimal traction speed command is output and sent to the traction drive system, and the coal mining machine is driven to run smoothly according to the optimal traction speed setpoint; During the cutting process, real-time feedback parameters are continuously monitored to ensure that the operating status is always within the safe constraint range; S7: Knowledge Fusion Update and Cross-Cycle Iteration; After the N+1th cut is completed, the coal and rock property identification results obtained through real-time correction in the current cutting cycle are fused and updated with the original coal and rock property spatial mapping information in the knowledge base; The updated coal and rock property spatial mapping information is rewritten into the knowledge base as the prior basis for the next cutting cycle; By repeatedly executing S1 to S7 in multiple cutting cycles, the coal and rock property information and traction control effect are continuously iterated and updated, realizing the cross-cycle rolling optimization of the coal mining machine traction speed and the adaptive evolution of the system. Thus, the system has the ability to continuously adapt and evolve with the advancement of the working face, thereby forming a cross-cycle rolling optimization mechanism for the coal mining machine traction speed.

[0026] As a preferred embodiment, in S1, the motion state parameters include traction speed, motor current, and power, etc.

[0027] As a preferred option, in S1, the temperature field information of the coal wall is collected by an infrared thermal imaging device installed at the front of the coal mining machine.

[0028] As a preferred option, in S2, the characteristic parameters include temperature gradient characteristic parameters, heat distribution characteristic parameters, etc.

[0029] As a preferred option, in S2, the coal and rock properties include information such as hardness, joints, and coal-rock interfaces.

[0030] As a preferred option, in S6, the feedback parameters include parameters such as traction load and motor current.

[0031] As a preferred option, in S5, the multi-objective constraint conditions of the multi-objective constraint optimization model include at least the traction system load constraint, motor current / power constraint, and cutting safety constraint.

[0032] This invention provides a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing. First, it simultaneously collects the coal face temperature field, the operating parameters of the coal mining machine, and location information to construct a multi-source sensing dataset. Compared to relying on a single sensor, temperature field information can detect changes in coal and rock in advance (thermal response precedes changes in mechanical load), providing richer and earlier input for subsequent identification and decision-making. Simultaneously, location information and temperature field are collected synchronously, laying a precise spatiotemporal benchmark for constructing a spatial mapping relationship between location and coal and rock properties. Second, it utilizes the thermal response characteristics of coal and rock friction during cutting to achieve non-contact, real-time identification of coal and rock properties, avoiding the wear and failure problems of traditional contact sensors. Since thermal response characteristic parameters have an inherent physical correlation with the physical properties of coal and rock (hardness, brittleness, coefficient of friction), the extraction of thermal response characteristic parameters makes the identification results more interpretable and reliable, rather than a black-box model output. Next, the coal and rock property identification results are spatiotemporally correlated with the location of the coal mining machine to construct a spatial mapping relationship between location and coal and rock properties. This achieves a digital, visual, and continuous representation of coal and rock distribution characteristics, providing map-like prior knowledge for subsequent cutting and a stable and reliable prior basis for subsequent traction control. Simultaneously, for the first time, coal and rock property information is stored in a structured form in a knowledge base, giving the system memory capabilities and breaking the traditional independent, memoryless control mode of each cut of the coal mining machine. Subsequently, during the N+1 cut, the prior mapping (coal and rock map) from the previous cut is invoked as a decision guide, while real-time temperature field sensing is used to monitor and verify actual coal and rock changes. This mechanism of prior knowledge combined with real-time dual-channel verification effectively overcomes the problem of inaccuracy caused by sudden changes in coal and rock (such as faults or scour zones) due to a single prior knowledge, and also avoids misjudgments caused by local thermal interference due to a single real-time sensing. Through mutual verification between prior knowledge and real-time data, the corrected coal and rock property identification results are output, achieving continuous improvement in sensing accuracy as the cutting process progresses. Furthermore, by constructing a multi-objective constrained optimization model, multiple mutually restrictive objectives such as cutting stability, load fluctuation suppression, energy consumption reduction, and cutting efficiency are simultaneously considered. Compared to traditional single-objective control methods (such as constant power or constant speed), this method can achieve a more balanced and intelligent traction speed decision-making process under complex coal and rock conditions. It can effectively suppress traction load fluctuations, reduce mechanical specific energy, and significantly improve the stability and energy efficiency of the cutting process. The optimal traction speed setpoint is calculated in real time based on the current (corrected) coal and rock properties, which helps to achieve refined control that adapts to different rock conditions and adjusts speed accordingly.Then, after the optimal speed command is sent to the traction system, real-time feedback parameters (such as motor current, vibration, and temperature rise) are continuously monitored to ensure that the operating status remains within safe constraints. This approach embodies the composite control concept of feedforward and feedback, effectively responding to changes in coal and rock while ensuring equipment safety. Simultaneously, this method fully considers mechanical inertia and response time, avoiding mechanical shock conditions caused by sudden speed changes. Finally, after each cut, the updated data obtained from real-time correction is integrated with the original information in the knowledge base and rewritten into the knowledge base. This ensures that the knowledge base can continuously update and optimize itself as the working face advances, giving the system adaptive evolution capabilities. By repeatedly executing S1 to S7 in multiple cutting cycles, cross-cycle rolling optimization of traction speed decisions is achieved. Each cut is both a response to the current coal and rock conditions and a learning process for the next cut. This mechanism allows the system to gradually adapt to long-term coal seam changes (such as gradual changes in coal quality along the working face) and continuously evolve adaptively with the cutting process, improving the stability and reliability of the coal mining machine under complex coal and rock conditions. Meanwhile, by employing a cross-cycle priori guidance mechanism, the reliance on fully real-time coal and rock identification and high-performance computing capabilities is reduced, improving the system's engineering applicability in complex underground environments. This invention focuses on the reciprocating cutting operation of a coal mining machine in a fully mechanized mining face. Using the single-cut cutting process as the basic control cycle unit, a cross-cycle closed-loop control mechanism is constructed, encompassing perception, identification, mapping, priori guidance and real-time correction, traction optimization, execution, and rolling updates. During the Nth cut, the system emphasizes the perception and spatial mapping of coal and rock properties. During the N+1th cut, the system makes traction control decisions based on the coal and rock property distribution information from the previous cut and continuously corrects for changes in coal and rock through continuous perception. This iterative optimization of the control effect across multiple cutting cycles enables the coal mining machine's traction control process to continuously learn and optimize, overcoming the limitations of traditional fixed parameters or manual intervention.

[0033] This method is simple to implement and has low implementation costs. It can achieve adaptive rolling optimization of the traction speed of the coal mining machine, and while taking into account cutting efficiency and energy consumption, it significantly improves cutting stability and intelligent decision-making level under complex geological conditions.

[0034] This invention also provides a cross-cycle control system for the traction speed of a coal mining machine based on temperature field sensing, used to implement a cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, comprising: The sensing and acquisition module is installed on the coal mining machine and is used to collect coal wall temperature field information, coal mining machine (traction speed, motor current, power, etc.) operating status parameters, and coal mining machine position information in the working face in a non-contact manner. A thermal response feature extraction module, which is connected to the sensing and acquisition module, is used to extract features from the coal wall temperature field information to obtain thermal response feature parameters; A coal and rock property identification module, which is connected to a thermal response feature extraction module, is used to identify and classify coal and rock properties based on thermal response feature parameters and output the coal and rock property identification results at the current location. The spatial mapping and knowledge base storage module is connected to the coal and rock property identification module. It is used to spatiotemporally associate the coal and rock property identification results with the coal mining machine location information, construct the location-coal and rock property spatial mapping relationship, and store the coal and rock property spatial mapping information in the knowledge base as prior information for subsequent cutting cycles. The prior knowledge retrieval and correction module is connected to both the knowledge base and the sensing and acquisition module. In subsequent cutting cycles, it retrieves the corresponding coal and rock property spatial mapping information generated in the previous cutting cycle from the knowledge base based on the real-time position of the coal mining machine, using it as the prior input for the coal and rock properties at the current position. Simultaneously, it is used to identify and monitor changes in coal and rock properties in real time based on the real-time acquired coal wall temperature field information, achieving mutual verification and correction between prior knowledge and real-time sensing, and outputting the corrected coal and rock property identification result. A multi-objective traction speed optimization module is connected to both the sensing and acquisition module and the prior call and correction module. Based on the corrected coal and rock property identification results, the module determines the optimal traction speed setpoint through a multi-objective optimization algorithm under multi-objective constraints. Simultaneously, it monitors the coal mining machine's operating feedback parameters in real time and ensures that the cutting process is always within the safe constraint range, forming a control closed loop. The traction control execution module is used to take the optimal traction speed setpoint as the control input, make a decision and output the optimal traction speed command to the traction drive system, so as to realize the adaptive adjustment of the traction speed of the coal mining machine.

[0035] As a preferred embodiment, a rolling update module is also included. The rolling update module is used to merge and update the coal and rock property identification results obtained by real-time correction in the current cutting cycle with the original coal and rock property spatial mapping information in the knowledge base, generate updated coal and rock property spatial mapping information and rewrite it into the knowledge base as a priori basis for the next cutting cycle, thereby realizing cross-cycle rolling update of coal and rock property information.

[0036] In this technical solution, by setting up a rolling update module, not only is the control effect of a single cutting cycle improved, but more importantly, the system is given the vitality of continuous evolution, so that the traction control of the coal mining machine can continuously improve itself as the working face advances, and finally achieves true adaptive and intelligent operation.

[0037] As a preferred embodiment, the sensing and acquisition module includes an infrared thermal imaging device, a multi-parameter monitoring integrated device, and a positioning monitoring device; the infrared thermal imaging device is used to acquire coal face temperature field information; the multi-parameter monitoring integrated device is used to acquire operating status parameters of the coal mining machine (traction speed, motor current, power, etc.); and the positioning monitoring device is used to acquire the position information of the coal mining machine in the working face.

[0038] In this invention, the sensing and acquisition module enables non-contact sensing of coal wall temperature, avoiding sensor wear issues caused by direct contact with coal and rock, thus improving system reliability and lifespan. Simultaneously, the synchronous acquisition of multi-dimensional data provides a rich data foundation for subsequent accurate identification and control. The thermal response feature extraction module transforms raw, complex temperature field data into thermal response feature parameters with clear physical meaning, such as temperature change rate and thermal diffusivity. This process achieves a crucial transformation from data to features, significantly reducing the complexity of subsequent identification. Furthermore, this feature extraction process effectively filters out environmental noise and interference, highlighting temperature response changes caused by differences in the physical properties of coal and rock, improving information effectiveness and robustness of identification. The coal and rock property identification module leverages the inherent differences in thermal conductivity, thermal diffusivity, and other thermophysical properties between coal and rock for convenient and accurate identification. Compared to indirect methods relying on consequential parameters such as vibration and current during cutting, temperature field-based identification is more predictive and accurate. Therefore, the coal and rock properties at the current location can be classified and identified in real time during the movement of the coal mining machine, providing an immediate basis for dynamic adjustments. Through the setting of spatial mapping and knowledge base storage modules, the identified coal and rock properties are innovatively correlated with the precise location of the coal mining machine in time and space, constructing a spatial mapping relationship between location and coal and rock properties. This transforms the information perceived during the cutting process into structured knowledge for storage, making each cutting operation a process of accumulating experience for the next cutting operation, achieving knowledge accumulation and reuse. Through the setting of prior knowledge retrieval and correction modules, the historical experience in the knowledge base (prior information from the previous cycle) is creatively combined with the real-time perceived information of the current cycle. This dual-source information verification mechanism greatly improves the reliability and accuracy of coal and rock property identification. Simultaneously, when there is a deviation between real-time perception and prior knowledge, the system can correct based on real-time perception. This can not only cope with sudden changes in local geological conditions but also reverse-verify and optimize the knowledge base, enabling the knowledge base to have self-learning and dynamic update capabilities. By configuring a multi-objective traction speed optimization module, the results of coal and rock property identification and correction can be used as input. Within a unified framework, multiple, even conflicting, objectives such as production efficiency (high traction speed), equipment safety (low load), and energy consumption are comprehensively considered. An optimization algorithm is used to find the optimal, rather than the fastest, traction speed. Based on this, the operation feedback of the coal mining machine is monitored in real time to ensure that any optimization decision will not cause the equipment to operate beyond its safety boundaries. This constitutes a complete control closed loop of perception-decision-execution-feedback, guaranteeing the safe and stable operation of the system. Through the configuration of the traction control execution module, the optimal speed setpoint obtained from the optimization algorithm can be accurately and quickly converted into control commands that the drive system can recognize, ensuring that the upper-level intelligent decisions are accurately implemented in the lower-level physical world.This invention takes temperature field sensing as an innovative entry point, and combines spatial knowledge graphs and multi-objective optimization technology to transform passive interception into active perception and pre-adaptation, single control into multi-objective collaborative optimization, and single interception into cross-cycle self-learning, thus constructing an intelligent control ecosystem that integrates prediction, memory, optimization, and correction.

[0039] Through the collaborative work of multiple modules, the system achieves cross-cycle decoupling and rolling optimization of coal and rock property perception and coal mining machine traction speed control, effectively improving the adaptability and operational stability of the coal mining machine in environments with limited computing power and complex communication conditions.

Claims

1. A cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, characterized in that... It includes the following steps: S1: Data acquisition and status perception; during the Nth cut, coal wall temperature field information is acquired; coal mining machine operating status parameters are acquired simultaneously, and the position information of the coal mining machine in the working face is obtained to construct a multi-source perception dataset for the cutting process; S2: Thermal response feature extraction and coal and rock property identification; Based on the collected coal wall temperature field information, feature parameters reflecting the thermal response characteristics of coal and rock are extracted, and the coal and rock property identification and classification are completed through the preset coal and rock property identification model to obtain the coal and rock property identification results at the current location; S3: Spatial mapping construction and knowledge storage; The coal and rock property identification results are spatiotemporally correlated with the coal mining machine location information to construct a spatial mapping relationship between location and coal and rock properties, which is used to characterize the distribution characteristics of coal and rock properties along the working face in the current cutting cycle; At the same time, the spatial mapping information of coal and rock properties is stored in the knowledge base as a priori basis for subsequent cutting cycles. S4: Prior knowledge retrieval and real-time correction; During the N+1th cut, based on the real-time position of the coal mining machine, the corresponding coal and rock property spatial mapping information formed by the previous cut cycle is retrieved from the knowledge base to obtain the corresponding coal and rock property at the current position as prior input; At the same time, coal wall temperature field information is continuously collected to identify and monitor the actual changes in coal and rock properties in real time, realizing mutual verification and correction between prior knowledge and real-time perception, and outputting the corrected coal and rock property identification result; S5: Multi-objective constraint index speed decision; After obtaining the corrected coal and rock property identification results, construct a multi-objective constraint optimization model; Under the premise of satisfying the multi-objective constraint conditions, with the goal of improving cutting stability, reducing traction load fluctuations, reducing cutting energy consumption, and taking cutting efficiency into account, determine the optimal traction speed setting value that matches the current coal and rock properties. S6: Traction Execution and Process Control; The optimal traction speed setpoint is used as the control input, the optimal traction speed command is output and sent to the traction drive system, and the coal mining machine is driven to run smoothly according to the optimal traction speed setpoint; During the cutting process, real-time feedback parameters are continuously monitored to ensure that the operating status is always within the safe constraint range; S7: Knowledge Fusion Update and Cross-Cycle Iteration; After the N+1th cut is completed, the coal and rock property identification results obtained through real-time correction in the current cutting cycle are fused and updated with the original coal and rock property spatial mapping information in the knowledge base; The updated coal and rock property spatial mapping information is rewritten into the knowledge base as the prior basis for the next cutting cycle; By repeatedly executing S1 to S7 in multiple cutting cycles, the cross-cycle rolling optimization of the coal mining machine traction speed and the adaptive evolution of the system are realized.

2. The method for cross-cycle control of coal mining machine traction speed based on temperature field sensing according to claim 1, characterized in that, In S1, the motion state parameters include traction speed, motor current, and power.

3. A cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing, as described in claim 1 or 2, is characterized in that... In S1, the temperature field information of the coal wall is collected by an infrared thermal imaging device installed at the front of the coal mining machine.

4. The method for cross-cycle control of coal mining machine traction speed based on temperature field sensing according to claim 3, characterized in that, In S2, the characteristic parameters include temperature gradient characteristic parameters and heat distribution characteristic parameters.

5. The method for cross-cycle control of coal mining machine traction speed based on temperature field sensing according to claim 1, characterized in that, In S2, coal and rock properties include hardness, joints, and coal-rock interface information.

6. The method for cross-cycle control of coal mining machine traction speed based on temperature field sensing according to claim 1, characterized in that, In S6, the feedback parameters include traction load and motor current parameters.

7. The method for cross-cycle control of coal mining machine traction speed based on temperature field sensing according to claim 1, characterized in that, In S5, the multi-objective constraints of the multi-objective constraint optimization model include at least the traction system load constraint, motor current / power constraint, and cutting safety constraint.

8. A cross-cycle control system for the traction speed of a coal mining machine based on temperature field sensing, used to implement the cross-cycle control method for the traction speed of a coal mining machine based on temperature field sensing as described in any one of claims 1 to 7, characterized in that, include: The sensing and acquisition module is installed on the coal mining machine and is used to collect coal wall temperature field information, coal mining machine operating status parameters, and coal mining machine position information in the working face in a non-contact manner. A thermal response feature extraction module, which is connected to the sensing and acquisition module, is used to extract features from the coal wall temperature field information to obtain thermal response feature parameters; A coal and rock property identification module, which is connected to a thermal response feature extraction module, is used to identify and classify coal and rock properties based on thermal response feature parameters and output the coal and rock property identification results at the current location. A spatial mapping and knowledge base storage module is connected to a coal and rock property identification module. It is used to spatiotemporally associate the coal and rock property identification results with the coal mining machine location information, construct a location-coal and rock property spatial mapping relationship, and store the coal and rock property spatial mapping information in a knowledge base. The prior knowledge retrieval and correction module is connected to both the knowledge base and the sensing and acquisition module. In subsequent cutting cycles, it retrieves the corresponding coal and rock property spatial mapping information generated in the previous cutting cycle from the knowledge base based on the real-time position of the coal mining machine, using it as the prior input for the coal and rock properties at the current position. Simultaneously, it is used to identify and monitor changes in coal and rock properties in real time based on the real-time acquired coal wall temperature field information, achieving mutual verification and correction between prior knowledge and real-time sensing, and outputting the corrected coal and rock property identification result. A multi-objective traction speed optimization module is connected to both the sensing and acquisition module and the prior call and correction module. Based on the corrected coal and rock property identification results, the module determines the optimal traction speed setpoint through a multi-objective optimization algorithm under multi-objective constraints. Simultaneously, it monitors the coal mining machine's operating feedback parameters in real time and ensures that the cutting process is always within the safe constraint range, forming a control closed loop. The traction control execution module is used to take the optimal traction speed setpoint as the control input, make a decision and output the optimal traction speed command to the traction drive system, so as to realize the adaptive adjustment of the traction speed of the coal mining machine.

9. A cross-cycle control system for the traction speed of a coal mining machine based on temperature field sensing, as described in claim 8, is characterized in that... Also includes: The rolling update module is used to merge and update the coal and rock property identification results obtained by real-time correction in the current cutting cycle with the original coal and rock property spatial mapping information in the knowledge base, generate updated coal and rock property spatial mapping information and rewrite it into the knowledge base as a priori basis for the next cutting cycle, so as to realize cross-cycle rolling update of coal and rock property information.

10. A cross-cycle control system for the traction speed of a coal mining machine based on temperature field sensing, as described in claim 8, is characterized in that... The sensing and acquisition module includes an infrared thermal imaging device, a multi-parameter monitoring integrated device, and a positioning monitoring device; the infrared thermal imaging device is used to acquire coal face temperature field information; the multi-parameter monitoring integrated device is used to acquire coal mining machine operating status parameters; and the positioning monitoring device is used to acquire the coal mining machine's position information in the working face.

Images