A method for straightening a scraper conveyor based on a rolling horizon control concept

Abstract

The application discloses a kind of based on rolling time domain control concept's scraper conveyor straightening method, it is solved that the existing scraper conveyor straightening method exists and cannot realize straightening situation because of not considering the dynamic change of coal seam floor form, cannot realize single cutter straightening and other problems because of information hysteresis;The scraper conveyor straightening method includes execution space, deduce space, prediction space;Execution space receives the final determination of scraper conveyor optimal push and roll strategy of prediction space, controls the push and roll operation of subsequent scraper conveyor by electro-hydraulic control system according to optimal push and roll strategy to reach the purpose of scraper conveyor straightening;While, working face real-time information is fed back to deduce space;Deduce space deduces working face real-time information, and deduce information is sent to prediction space;Prediction space receives the deduce information of deduce space and carries out analog prediction, finally determines the optimal push and roll strategy of scraper conveyor.

Description

Technical Field

[0001] This invention relates to the field of coal mining process control in underground fully mechanized mining faces, specifically a method for straightening scraper conveyors based on the concept of rolling time-domain control. Background Technology

[0002] Scraper conveyors are a crucial component of fully mechanized coal mining faces. In the coordinated operation of three mining machines (scraper, conveyor, and machine), the scraper conveyor not only serves as the running track for the mining machine but also plays a vital role in transporting the coal cut and harvested by the machine. Furthermore, the scraper conveyor is connected to numerous hydraulic supports, enabling push-conveyor operations. Adjusting the position and orientation of the scraper conveyor according to its straightness prevents material blockage during transport, thereby improving coal mine production efficiency. Simultaneously, straightening the scraper conveyor significantly reduces equipment failure and wear, thus greatly minimizing equipment maintenance time, improving the safety of the working environment, and reducing the labor intensity and accident risk for workers.

[0003] Patent document CN201810204691.5 discloses a method for straightening a scraper conveyor based on the effective pushing stroke of hydraulic supports. Each hydraulic support is equipped with relevant sensors, and the controllers of all hydraulic supports are connected together. One hydraulic support is randomly selected, and its effective pushing displacement is obtained using pressure and displacement sensors. This effective pushing displacement is used as an expected value. Based on this expected value, other hydraulic supports are controlled to perform pushing operations. The effective pushing displacement obtained after pushing is compared with the expected value, and adjustments are made based on this comparison result. Finally, the deviation of each hydraulic support after pushing is reduced, thereby ensuring the straightness of the scraper conveyor.

[0004] Patent document CN201510379025.1 discloses an automatic straightening device and method for the body of a scraper conveyor in a fully mechanized mining face. An elastic rod is installed between two adjacent hydraulic supports, and an angle sensor is installed between the elastic rod and the hydraulic support. An elastic connector is installed between two adjacent sections of the scraper conveyor's central trough, and a strain sensor capable of temperature compensation is installed in the elastic connector. An information processing system connects the angle sensor and the strain sensor, and a data transmission module establishes communication between the information processing system and the electro-hydraulic control system. The positioning of the hydraulic supports and the straightening operation of the central trough of the scraper conveyor are based on the voltage signal received by the electro-hydraulic control system, which then controls the hydraulic supports and the scraper conveyor to perform relevant actions according to the actual situation.

[0005] Patent document CN202210319338.8 provides a method for straightening a scraper conveyor, an electronic device, and a storage medium. The method involves acquiring the trajectory of a coal mining machine cutting the coal face, calculating the corresponding scraper conveyor trajectory, and then comparing the corresponding scraper conveyor trajectory with a reference straight line perpendicular to the working face's advancing direction. Based on the comparison result between the scraper conveyor trajectory corresponding to the previous coal face cutting and the reference straight line, the method calculates the displacement compensation value of each push point on the scraper conveyor trajectory corresponding to the current coal face cutting. In the current coal face cutting, each push point is compensated based on its displacement compensation value. Using the scraper conveyor as a reference, the pushing distance of the subsequent scraper conveyor is determined through the previous scraper conveyor trajectory, thereby performing a straightening operation.

[0006] However, the drawback of the above method is:

[0007] (1) The effective pushing displacement obtained after pushing is compared with the expected value. Only the pushing process of the scraper conveyor is considered in the pushing process. In the actual pushing process, due to the change of the coal seam floor, the middle trough of the scraper conveyor may not be able to complete the pushing process, which does not conform to the actual pushing situation.

[0008] (2) Under actual mine conditions, the coal seam floor is unpredictable, especially during the cutting process. Simply considering the use of electro-hydraulic control system to adjust the scraper conveyor and hydraulic support may result in situations where certain stages of the pushing process cannot be achieved due to changes in the floor shape.

[0009] (3) Obtaining the trajectory of the scraper conveyor requires inverting the information after the coal mining machine cuts one cut, and then straightening the scraper conveyor based on the trajectory information. Due to the lag in information, single-cut straightening cannot be achieved. Summary of the Invention

[0010] To address the problems of existing scraper conveyor straightening methods, such as the inability to achieve straightening due to the lack of consideration for dynamic changes in the coal seam floor morphology and the inability to achieve single-blade straightening due to information lag, this invention provides a scraper conveyor straightening method based on the concept of rolling time-domain control.

[0011] The present invention is achieved by the following technical solution: a method for straightening a scraper conveyor based on the concept of rolling time domain control, comprising the following spaces: an execution space including a feedback control model, a derivation space including a working face information processing model, and a prediction space including a coupled base plate update model, a baseline prediction model, a spatial difference feedback model, and a control quantity optimization model;

[0012] The feedback control model in the execution space receives the optimal pushing strategy of the scraper conveyor finally determined by the prediction space. Based on the optimal pushing strategy, the electro-hydraulic control system controls the subsequent pushing operation of the scraper conveyor to achieve the purpose of straightening the scraper conveyor. At the same time, the real-time information of the working face is fed back to the working face information processing model in the simulation space.

[0013] The working face information processing model in the simulation space performs inference and inversion (referred to as inference) on the real-time information of the working face fed back by the execution space feedback control model after it runs, and sends the inference information (i.e., the inference and inversion results) to the prediction space; the real-time information of the working face includes the position and posture information of the three fully mechanized mining machines, the cutting information of the coal mining machine, and the information of the coal seam floor, etc.

[0014] The prediction space receives the simulation information from the deduction space to perform simulation prediction and finally determine the optimal pushing strategy of the scraper conveyor; wherein, the coupled floor update model constructs a virtual known coal seam floor based on the simulation information, establishes the coupling relationship between the fully mechanized mining equipment and the coal seam floor, and finally predicts and updates the coal seam floor model of the area to be mined.

[0015] The baseline prediction model is based on the aforementioned coupled base plate update model. It simulates the pushing process of the scraper conveyor and predicts the baseline of the scraper conveyor after pushing, providing a calculation basis for the subsequent spatial difference feedback model and control quantity optimization model.

[0016] The spatial difference feedback model provides correction data for the equipment pose adjustment of the subsequent control quantity optimization model; the spatial difference refers to the actual pose information of the scraper conveyor after the push-slide in the current (i.e., during this cut) and the spatial difference between the baseline prediction model and the baseline of the scraper conveyor after the push-slide predicted in the previous (i.e., during the previous cut).

[0017] The control quantity optimization model obtains the optimal push-pull strategy based on the coupled base plate update model, baseline prediction model, and spatial difference feedback model, and provides it to the feedback control model of the execution space.

[0018] The specific process by which the working face information processing model acquires the position and pose information of the three fully mechanized mining machines is as follows:

[0019] Step 101: Hydraulic Support Group Pose Information Acquisition Steps: Install a 3D LiDAR at the middle of the coal mining machine body and scan the hydraulic support group during the coal mining machine cutting process; take the pin shaft connecting the base and the push cylinder on the hydraulic support, and the vertex position of the connection between the top beam and the side guard plate as feature parts, acquire the point cloud data of the two feature parts, and perform filtering, segmentation, and registration operations; calculate the relatively accurate point cloud data to finally obtain the pose information of the hydraulic support group;

[0020] Step 102, Scraper Conveyor Position Information Acquisition Steps: Acquire data from the strapdown inertial navigation system on the coal mining machine; eliminate the accumulated error of the strapdown inertial navigation system using the extended Kalman filter method to obtain accurate coal mining machine position information; based on the positional relationship between the scraper conveyor and the coal mining machine, invert the scraper conveyor trajectory using the coal mining machine position information; select key points on the middle trough of each section of the scraper conveyor and use the key points to calculate the scraper conveyor position information; there are no special requirements for the selection of the key points, generally the center point of the middle trough of each section of the scraper conveyor is selected;

[0021] Step 103, Data Processing Steps: Eliminate outliers in the pose information obtained in the above steps; quantify and store the information data to facilitate subsequent information deduction and update operations;

[0022] Furthermore, the working face information processing model can acquire coal mining machine cutting information and coal seam floor information using existing known technologies;

[0023] The specific construction process of the coupled base plate update model is as follows:

[0024] Step 201: Construct a virtual coal seam floor: Based on the known coal seam floor information obtained from the working face information processing model, construct a virtual coal seam floor using the Mesh component in Unity3D software;

[0025] Step 202: Construct a coupling relationship model between the virtual coal seam floor and the equipment model: Import the equipment model into Unity3D software and set relevant rigid body components for the equipment model; adjust the position parameters of the equipment model so that the equipment model and the virtual coal seam floor can fit together fully, thus completing the construction of the coupling relationship model between the virtual coal seam floor and the equipment model; the equipment includes a coal mining machine, a hydraulic support group, and a scraper conveyor;

[0026] Step 203: Predict the coal seam floor correction model corresponding to the push of the scraper conveyor: Based on the coal cutting information (the cutting information of the rear drum during coal cutting) obtained from the working face information processing model and the coal drop information during the previous cutting process of the coal mining machine, the deep LSTM (Long Short-Term Memory) neural network method is applied to predict the coal seam floor correction model corresponding to the push of the scraper conveyor; the deep LSTM (Long Short-Term Memory) neural network method is a known existing technology and is an algorithmic mathematical model that imitates the behavioral characteristics of the animal nervous system, used for distributed parallel information processing.

[0027] Step 204: Predict the physical behavior-based coal seam floor update model: Based on the coupling relationship model in Step 202 and the coal seam floor correction model in Step 203, set the scraper conveyor pushing amount, run Unity3D software to simulate the scraper conveyor pushing operation, apply the SURF algorithm to extract the feature point information of the coal seam floor corresponding to the pushing section after the scraper conveyor pushing, reconstruct the coal seam floor corresponding to the pushing section after the scraper conveyor pushing, and obtain the coal seam floor reconstruction model; based on the coal seam floor reconstruction model, analyze the damage of the scraper conveyor pushing behavior to the two situations of pits and loose coal piles in the coal seam floor (coal seam floor). The coal seam can be broadly categorized into two types: pitted and loose coal piles. A stress analysis was performed, and the failure criteria for the coal seam floor after the scraper conveyor pushes the coal seam were derived based on the Mohr-Coulomb criterion. Based on these failure criteria, a numerical simulation analysis of the coal seam floor reconstruction model was conducted using FLAC3D software. The maximum depth and location of damage caused to the coal seam floor by the support pressure along the scraper conveyor pushing direction and the lateral direction under the stress model were calculated. Finally, based on the coal seam floor reconstruction model and the results of the maximum damage depth and location, a physical behavior-based coal seam floor update model was constructed using the Mesh component in Unity3D software.

[0028] The specific construction process of the baseline prediction model is as follows:

[0029] Step 301: Based on the first-level coordinate system, represent the position information of the key points of the middle trough of each section of the scraper conveyor (i.e., before the pusher); there are no special requirements for the selection of the key points of the middle trough, and the center point of the middle trough of each section of the scraper conveyor is generally selected.

[0030] Step 302: Compare the coordinate information of each section of the middle trough in the pushing direction of the scraper conveyor, determine the last section of the middle trough, and record its sequence number i;

[0031] Step 303: Based on the coal seam floor update model, establish a parallel system in Unity3D software to simulate the pushing process of the scraper conveyor and obtain the position information after the last remaining middle trough (i.e. the i-th middle trough) has been pushed to its full stroke (i.e., the maximum stroke that the pushing cylinder of the space electro-hydraulic control system can push the middle trough).

[0032] Step 304: Select n position points evenly throughout the entire pushing stroke of the last section of the middle groove, label the n position points from 1 to n in the direction from the start of pushing to the end of pushing, and obtain the position information of each position point;

[0033] Step 305: Using the position line corresponding to position point n (the position line is a reference straight line passing through position point n and perpendicular to the pushing direction of the scraper conveyor on the working surface) as the end point of the pushing stroke, simulate the pushing process of the remaining middle troughs of the scraper conveyor in the aforementioned parallel system, and determine whether the key points of the other middle troughs can all reach the position line corresponding to position point n. If they can all reach it, then the position line is the predicted baseline of the scraper conveyor after pushing. If they cannot all reach it, then select the position line corresponding to the previous position point as the end point of the pushing stroke, and repeat the above simulation and judgment process until the key points of the remaining middle troughs of the scraper conveyor can all reach the end point of the pushing stroke.

[0034] The spatial difference feedback model consists of three parts: feedforward space, feedback space, and correction mechanism. The specific construction process is as follows:

[0035] Step 401, Feedforward Space: Based on the aforementioned baseline prediction model, the process of predicting the baseline after the scraper conveyor pushes the conveyor is predicted. The prediction patterns in the prediction process are extracted, the possible situations are estimated, and corresponding measures are taken to eliminate possible deviations in advance. For example, when the slope of the coal seam floor is relatively large, a phenomenon called upward sliding will occur. Or, in special terrain, the hydraulic support may not reach the ideal position, resulting in unsmooth pushing. The prediction patterns are summarized. During the pushing process, the possible situations can be summarized by looking at the predicted coal seam floor. Then, appropriate measures can be selected to eliminate possible deviations.

[0036] Step 402, Feedback Space: Based on the baseline of the scraper conveyor after the push (i.e., the pose information of the scraper conveyor after the push is predicted before the push) and the actual pose information of the scraper conveyor after the push, the spatial difference is obtained by calculation.

[0037] Step 403, Correction Mechanism: Combining the data information obtained from the feedforward space and the feedback space, the correlation coefficient is applied to the feedforward space prediction process to obtain the corrected prediction result; continuously monitor the feedback data and adjust it according to the real-time feedback information; continuously iterate and optimize the correction mechanism based on the feedback results of actual operation.

[0038] The specific construction process of the control quantity optimization model is as follows:

[0039] Step 501: Obtain the initial push control quantity: Obtain the pose information of the current scraper conveyor push section from the working face information processing model (i.e., the actual pose information of the scraper conveyor after the previous push and before the current push, referred to as the actual pose information before the push). Predict the baseline of the scraper conveyor after the current push using the baseline prediction model (i.e., the predicted baseline of the scraper conveyor after the push before the push, referred to as the baseline of the scraper conveyor before the push). Then, based on the spatial difference obtained from the spatial difference feedback model, correct the predicted baseline of the scraper conveyor after the current push. Obtain the initial push control quantity by calculating the difference between the corrected scraper conveyor baseline and the pose information of the current scraper conveyor push section.

[0040] Step 502, Floor Segmentation: Based on the initial calculated pusher control quantity and the coal seam floor update model based on physical behavior predicted in Step 204, obtain the floor data corresponding to the pusher section of the scraper conveyor, and the pusher trajectory information of the middle trough of each section of the scraper conveyor pusher; establish the floor function Z = F(x,y,z) corresponding to the pusher trajectory of the middle trough of each section of the scraper conveyor pusher based on the floor data, and find the extreme points of the floor function, letting... The process involves solving for points that may become extreme points and determining whether they are indeed extreme points. Ultimately, the extreme points of the coal seam floor corresponding to the pushing trajectory of the middle trough in each section of the scraper conveyor's pushing section are determined. Based on these extreme points, the coal seam floor in the pushing direction of the scraper conveyor's pushing section is segmented. Simultaneously, based on the corresponding extreme points, an optimized discretization method is used to segment the pushing trajectory of the middle trough in each section of the scraper conveyor's pushing section. This optimized discretization method is a known existing method that involves linking independent variables with target variables for discretization. In this invention, the target variable is the coordinate change ΔX of the pushing control quantity in the pushing direction, and the independent variable is the X-coordinate value of the coal seam floor.

[0041] Step 503: Simulate and solve the real-time pushing position of the middle trough of each section of the scraper conveyor in Unity3D software: Establish a coordinate system based on the pushing mechanism between the scraper conveyor and the hydraulic support. With the pushing direction of the scraper conveyor as the X-axis, select two points on the side of the scraper conveyor with the pushing lug on each section of the middle trough as key points, and obtain the pose information of the two key points to obtain a vector information corresponding to the middle trough; perform deduction based on the floating connection mechanism model in the pushing mechanism to calculate the coordinates of the contact point between the connecting head pin of the floating connection mechanism and the pushing lug in the X-axis direction during the pushing process; calculate the position information of the contact point based on the above vector information; finally, calculate the real-time pushing position of each section of the middle trough of the scraper conveyor based on the equipment posture information obtained from the coal seam floor update model based on physical behavior and the working face information processing model predicted in Step 204.

[0042] Step 504: Simulate the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software: Calculate the number of trough sections in the middle of the S-shaped curved section using the existing calculation method for the length of the curved section, and determine the number of sections in the front and rear curved sections (i.e., separate the front and rear curved sections). Establish the parent-child relationship of the middle troughs in Unity3D software and set a limit - i.e., the maximum curvature of the middle trough of each section. Then calculate the pushing amount required for the middle trough of each section of the corresponding shape of the scraper conveyor, and execute the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software.

[0043] Step 505, Control Quantity Screening Step: Based on the pushing simulation performed in Unity3D software in the above steps, output the segment information of each section of the scraper conveyor's middle trough that cannot be pushed smoothly during the entire pushing process; and determine whether there is a segment of the middle trough in each section of the scraper conveyor's pushing section that cannot be pushed smoothly at each instant during the pushing process. If any segment cannot be pushed smoothly, the corresponding middle trough pushing control quantity information is eliminated until the segment information of each section of the scraper conveyor's pushing section that can be pushed smoothly, as well as the corresponding middle trough pushing control quantity information, are obtained.

[0044] Step 506: Construct the cost function and find the theoretical minimum: Based on the segmented selection of base plate feature points in step 502, the cost function is constructed as follows:

[0045]

[0046] Where x(t) is the current position of the scraper conveyor; x t `arget` represents the position of the starting point of the corresponding segment on the base plate (i.e., the dividing point between two segments); `t` represents the time taken for the scraper conveyor to push the conveyor across the corresponding segment on the base plate; `u(t)` represents the distance taken for the scraper conveyor to push the conveyor across the corresponding segment on the base plate. 2 α represents the force exerted by the hydraulic cylinder during the pushing process; α and β represent the proportions of positional cost and behavioral cost. This method focuses on the control variables during the process, so α = 0.7 and β = 0.3.

[0047] Based on the initial calculation of the push-pull control quantity in step 501, the theoretical minimum value is calculated according to the cost function listed in equation (1);

[0048] Step 507: Selection of Optimal Push Control Quantity: Based on the push control quantity of the central trough obtained in Step 505 and the cost function constructed in Step 506, calculate the cost function under different pushing distances and pushing forces, and compare it with the theoretical minimum value to determine whether the minimum value can be reached (i.e., less than or equal to the minimum value). If the minimum value is reached, directly derive the corresponding push control quantity of the central trough as the optimal push control quantity. If the minimum value is not reached, re-segment according to the extreme point in Step 502. Based on the extreme point, adjust the segmentation point position and translate it towards the push start direction to complete the re-segmentation. Based on the re-segmentation of the coal seam floor, adjust the pushing distance and pushing force of the scraper conveyor push section, and repeat the above simulation and judgment process in Unity3D software. Through continuous adjustment and optimization, the cost function reaches the minimum value.

[0049] Step 508: Formulate the optimal push-pull strategy: Formulate the optimal push-pull strategy based on the optimal push-pull control quantity obtained in step 507; and transmit the optimal push-pull strategy to the feedback control model of the execution space.

[0050] The design concept of the method described in this invention is to immediately perform simulations and predictions after the coal mining machine has passed several hydraulic supports to obtain information such as the new shape of the coal seam floor, and then determine the ideal position information of the scraper conveyor. This allows the scraper conveyor to be controlled to reach the designated position, completing all subsequent operations within the same cut, without information lag, which conforms to the rolling time-domain control concept. The rolling time-domain control is a control system concept based on real-time parameter adjustment, which can dynamically adjust control parameters according to the real-time operating status of the system and external conditions to achieve better performance and robustness.

[0051] Compared with existing technologies, unlike existing scraper conveyor straightening methods that predict after the coal mining machine has completed one cut, the method of this invention can perform the prediction operation after the coal mining machine has passed several supports and advance according to the obtained prediction position. There is no information lag, and it can achieve straightening within a single cut. In addition, considering that the scraper conveyor will cause a certain degree of damage to the coal seam floor during the pushing stroke, the method of this invention adds a physical behavior template to the previous floor information correction when updating the floor. This more realistically reflects the changes in the coal seam floor during the actual pushing process, and can effectively avoid the situation where straightening cannot be achieved due to the dynamic changes in the shape of the coal seam floor during actual straightening control. At the same time, considering the situation where some intermediate troughs cannot reach the pushing stroke during the pushing process, the pushing process is simulated and screened to obtain the optimal control quantity that meets the pushing stroke, which is closer to the actual situation and facilitates the straightening of the scraper conveyor.

[0052] In summary, the method of the present invention has no information lag and can achieve single-blade straightening; the process control is practical; the coal seam floor update takes into account the physical damage to the floor caused by the scraper conveyor pushing the conveyor; the method of the present invention can efficiently achieve scraper conveyor straightening based on actual conditions. Attached Figure Description

[0053] Figure 1 This is a block diagram illustrating the working principle of the method described in this invention.

[0054] Figure 2 A block diagram illustrating the working principle of the coupled base plate update model;

[0055] Figure 3 This is a block diagram illustrating the working principle of the baseline prediction model.

[0056] Figure 4 This is a block diagram illustrating the working principle of the control quantity optimization model.

[0057] Figure 5 This is a schematic diagram showing the position of the contact point between the connector of the floating connection mechanism and the connecting hole of the middle groove pushing lug during the pushing process;

[0058] In the diagram: 1, 2 - key points; 3 - floating connection mechanism connector pin; 4 - contact point; 5 - central groove; 6 - push-moving ear hole. Detailed Implementation

[0059] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0060] A method for straightening a scraper conveyor based on the concept of rolling time-domain control, such as... Figure 1 As shown, the space includes the following: the execution space containing the feedback control model, the derivation space containing the working face information processing model, and the prediction space containing the coupled base plate update model, the baseline prediction model, the spatial difference feedback model, and the control quantity optimization model.

[0061] The feedback control model in the execution space receives the optimal pushing strategy of the scraper conveyor finally determined by the prediction space. Based on the optimal pushing strategy, the electro-hydraulic control system controls the subsequent pushing operation of the scraper conveyor to achieve the purpose of straightening the scraper conveyor. At the same time, the real-time information of the working face is fed back to the working face information processing model in the simulation space.

[0062] The working face information processing model in the simulation space performs inference and inversion (referred to as inference) on the real-time information of the working face fed back by the execution space feedback control model after it runs, and sends the inference information (i.e., the inference and inversion results) to the prediction space; the real-time information of the working face includes the position and posture information of the three fully mechanized mining machines, the cutting information of the coal mining machine, and the information of the coal seam floor, etc.

[0063] The prediction space receives the simulation information from the deduction space to perform simulation prediction and finally determine the optimal pushing strategy of the scraper conveyor; wherein, the coupled floor update model constructs a virtual known coal seam floor based on the simulation information, establishes the coupling relationship between the fully mechanized mining equipment and the coal seam floor, and finally predicts and updates the coal seam floor model of the area to be mined.

[0064] The baseline prediction model is based on the aforementioned coupled base plate update model. It simulates the pushing process of the scraper conveyor and predicts the baseline of the scraper conveyor after pushing, providing a calculation basis for the subsequent spatial difference feedback model and control quantity optimization model.

[0065] The spatial difference feedback model provides correction data for the equipment pose adjustment of the subsequent control quantity optimization model; the spatial difference refers to the actual pose information of the scraper conveyor after the push-slide in the current (i.e., during this cut) and the spatial difference between the baseline prediction model and the baseline of the scraper conveyor after the push-slide predicted in the previous (i.e., during the previous cut).

[0066] The control quantity optimization model obtains the optimal push-pull strategy based on the coupled base plate update model, baseline prediction model, and spatial difference feedback model, and provides it to the feedback control model of the execution space.

[0067] The specific process by which the working face information processing model acquires the position and pose information of the three fully mechanized mining machines is as follows:

[0068] Step 101: Hydraulic Support Group Pose Information Acquisition Steps: Install a 3D LiDAR at the middle of the coal mining machine body and scan the hydraulic support group during the coal mining machine cutting process; take the pin shaft connecting the base and the push cylinder on the hydraulic support, and the vertex position of the connection between the top beam and the side guard plate as feature parts, acquire the point cloud data of the two feature parts, and perform filtering, segmentation, and registration operations; calculate the relatively accurate point cloud data to finally obtain the pose information of the hydraulic support group;

[0069] Step 102, Scraper Conveyor Position Information Acquisition Steps: Acquire data from the strapdown inertial navigation system on the coal mining machine; eliminate the accumulated error of the strapdown inertial navigation system using the extended Kalman filter method to obtain accurate coal mining machine position information; based on the positional relationship between the scraper conveyor and the coal mining machine, invert the scraper conveyor trajectory using the coal mining machine position information; select key points on the middle trough of each section of the scraper conveyor and use the key points to calculate the scraper conveyor position information; there are no special requirements for the selection of the key points, generally the center point of the middle trough of each section of the scraper conveyor is selected;

[0070] Step 103, Data Processing Steps: Eliminate outliers in the pose information obtained in the above steps; quantify and store the information data to facilitate subsequent information deduction and update operations;

[0071] Furthermore, the working face information processing model can acquire coal mining machine cutting information and coal seam floor information using existing known technologies;

[0072] like Figure 2 As shown, the specific construction process of the coupled base plate update model is as follows:

[0073] Step 201: Construct a virtual coal seam floor: Based on the known coal seam floor information obtained from the working face information processing model, construct a virtual coal seam floor using the Mesh component in Unity3D software;

[0074] Step 202: Construct a coupling relationship model between the virtual coal seam floor and the equipment model: Import the equipment model into Unity3D software and set relevant rigid body components for the equipment model; adjust the position parameters of the equipment model so that the equipment model and the virtual coal seam floor can fit together fully, thus completing the construction of the coupling relationship model between the virtual coal seam floor and the equipment model; the equipment includes a coal mining machine, a hydraulic support group, and a scraper conveyor;

[0075] Step 203: Predict the coal seam floor correction model corresponding to the push of the scraper conveyor: Based on the coal cutting information (the cutting information of the rear drum during coal cutting) obtained from the working face information processing model and the coal drop information during the previous cutting process of the coal mining machine, the deep LSTM (Long Short-Term Memory) neural network method is applied to predict the coal seam floor correction model corresponding to the push of the scraper conveyor; the deep LSTM (Long Short-Term Memory) neural network method is a known existing technology and is an algorithmic mathematical model that imitates the behavioral characteristics of the animal nervous system, used for distributed parallel information processing.

[0076] In this embodiment, the specific process of using the deep LSTM neural network method to predict the coal seam floor correction model corresponding to the push conveyor is as follows: The input information of the coal mining machine's rear drum cutting at time t and the output coal seam floor update information of the LSTM neural unit at the previous time are simultaneously fed into the input gate, forget gate, and output gate of the LSTM neural unit. The weights of the three gates are calculated to obtain the gate values. The memory state of the LSTM neural unit is modified using the values ​​of the input gate, forget gate, and output gate, and the final output of the neural network is formed through an activation function. Data points at equal intervals are collected and normalized. x ij It is the cutting roller height value at the j-th sampling point in the i-th cut; x min With x max These are the minimum and maximum values ​​in the cutting drum height data, respectively; the data is divided into training and testing sets; the Adam algorithm is used to update the training set data; an appropriate range of hidden layer neurons is set, and appropriate hyperparameters are selected; the trained model is tested using the aforementioned testing set, and the error is recorded, thereby obtaining a more accurate and realistic coal seam floor correction model.

[0077] Step 204: Predict the physical behavior-based coal seam floor update model: Based on the coupling relationship model in Step 202 and the coal seam floor correction model in Step 203, set the scraper conveyor pushing amount, run Unity3D software to simulate the scraper conveyor pushing operation, apply the SURF algorithm to extract the feature point information of the coal seam floor corresponding to the pushing section after the scraper conveyor pushing, reconstruct the coal seam floor corresponding to the pushing section after the scraper conveyor pushing, and obtain the coal seam floor reconstruction model; based on the coal seam floor reconstruction model, analyze the damage of the scraper conveyor pushing behavior to the two situations of pits and loose coal piles in the coal seam floor (coal seam floor). The coal seam can be broadly categorized into two types: pitted and loose coal piles. A stress analysis was performed, and the failure criteria for the coal seam floor after the scraper conveyor pushes the coal seam were derived based on the Mohr-Coulomb criterion. Based on these failure criteria, a numerical simulation analysis of the coal seam floor reconstruction model was conducted using FLAC3D software. The maximum depth and location of damage caused to the coal seam floor by the support pressure along the scraper conveyor pushing direction and the lateral direction under the stress model were calculated. Finally, based on the coal seam floor reconstruction model and the results of the maximum damage depth and location, a physical behavior-based coal seam floor update model was constructed using the Mesh component in Unity3D software.

[0078] In this embodiment, the specific process of extracting feature point information of the coal seam floor corresponding to the pushing section of the scraper conveyor using the SURF algorithm is as follows: Based on the virtual coal seam floor constructed using the Mesh component in Unity3D software, and the coupling relationship model between the virtual coal seam floor and the equipment model, the coal seam floor is image-processed. Key feature points in the image are extracted using DoG (Difference of Gaussian) and LoG (Laplace Transform of Gaussian). In SURF, a box filter is used to approximate the LoG to obtain an integral image. Let an image be f(x,y), and the Hessian matrix is ​​used to extract feature points. The corresponding Hessian matrix is:

[0079]

[0080] Gaussian filtering is applied to the processed image for noise reduction. An image pyramid is constructed for multi-dimensional description. After obtaining the location of the points of interest and constructing the image pyramid, feature points are located within the points of interest. First, an appropriate threshold needs to be selected to retain the points with the strongest response among the points of interest; the larger the threshold, the more feature points are retained. Then, non-maximum suppression is used to compare the pixels. Finally, cubic linear interpolation is performed on the selected key points to obtain stable feature points. Based on these feature points, the coal seam floor corresponding to the pusher conveyor is reconstructed to obtain a coal seam floor reconstruction model.

[0081] like Figure 3 As shown, the specific construction process of the baseline prediction model is as follows:

[0082] Step 301: Based on the first-level coordinate system, represent the position information of the key points of the middle trough of each section of the scraper conveyor (i.e., before the pusher); there are no special requirements for the selection of the key points of the middle trough, and the center point of the middle trough of each section of the scraper conveyor is generally selected.

[0083] Step 302: Compare the coordinate information of each section of the middle trough in the pushing direction of the scraper conveyor, determine the last section of the middle trough, and record its sequence number i;

[0084] Step 303: Based on the coal seam floor update model, establish a parallel system in Unity3D software to simulate the pushing process of the scraper conveyor and obtain the position information after the last remaining middle trough (i.e. the i-th middle trough) has been pushed to its full stroke (i.e., the maximum stroke that the pushing cylinder of the space electro-hydraulic control system can push the middle trough).

[0085] Step 304: Select n position points evenly throughout the entire pushing stroke of the last section of the middle groove, label the n position points from 1 to n in the direction from the start of pushing to the end of pushing, and obtain the position information of each position point;

[0086] Step 305: Using the position line corresponding to position point n (the position line is a reference straight line passing through position point n and perpendicular to the pushing direction of the scraper conveyor on the working surface) as the end point of the pushing stroke, simulate the pushing process of the remaining middle troughs of the scraper conveyor in the aforementioned parallel system, and determine whether the key points of the other middle troughs can all reach the position line corresponding to position point n. If they can all reach it, then the position line is the predicted baseline of the scraper conveyor after pushing. If they cannot all reach it, then select the position line corresponding to the previous position point as the end point of the pushing stroke, and repeat the above simulation and judgment process until the key points of the remaining middle troughs of the scraper conveyor can all reach the end point of the pushing stroke.

[0087] The spatial difference feedback model consists of three parts: feedforward space, feedback space, and correction mechanism. The specific construction process is as follows:

[0088] Step 401, Feedforward Space: Based on the aforementioned baseline prediction model, the process of predicting the baseline after the scraper conveyor pushes the conveyor is predicted. The prediction patterns in the prediction process are extracted, the possible situations are estimated, and corresponding measures are taken to eliminate possible deviations in advance. For example, when the slope of the coal seam floor is relatively large, a phenomenon called upward sliding will occur. Or, in special terrain, the hydraulic support may not reach the ideal position, resulting in unsmooth pushing. The prediction patterns are summarized. During the pushing process, the possible situations can be summarized by looking at the predicted coal seam floor. Then, appropriate measures can be selected to eliminate possible deviations.

[0089] Step 402, Feedback Space: Based on the baseline of the scraper conveyor after the push (i.e., the pose information of the scraper conveyor after the push is predicted before the push) and the actual pose information of the scraper conveyor after the push, the spatial difference is obtained by calculation.

[0090] In this embodiment, quaternions are used to represent the baseline p of the scraper conveyor after the push-pull, predicted by the aforementioned baseline prediction model, and the actual pose information q of the scraper conveyor after the push-pull; using Q = q * p -1 Converting the above quaternions into axis angle form, the result is the spatial difference between the predicted baseline of the scraper conveyor after the push and the actual pose information of the scraper conveyor after the push.

[0091] Step 403, Correction Mechanism: Combining the data information obtained from the feedforward space and the feedback space, the correlation coefficient is applied to the feedforward space prediction process to obtain the corrected prediction result; continuously monitor the feedback data and adjust it according to the real-time feedback information; continuously iterate and optimize the correction mechanism based on the feedback results of actual operation.

[0092] In this embodiment, the Pearson correlation coefficient is used to assess linear correlation, and the prediction results are corrected based on the value of the Pearson correlation coefficient. If the correlation coefficient is close to 1, it indicates that the prediction results are highly correlated with the actual feedback data, and the original prediction results can be maintained. If the correlation coefficient is close to -1, it indicates that the prediction results are highly negatively correlated with the actual feedback data, and the prediction results need to be reversed. According to the degree of difference between the actual data and the prediction results, proportional adjustments or weighted adjustments are made to reduce the error. If the correlation coefficient is close to 0, it indicates that there is no linear relationship between the prediction results and the actual feedback data. In this case, a multinomial regression model is used to introduce the higher powers of the variables into the regression model to obtain their nonlinear relationship.

[0093] like Figure 4 As shown, the specific construction process of the control quantity optimization model is as follows:

[0094] Step 501: Obtain the initial push control quantity: Obtain the pose information of the current scraper conveyor push section from the working face information processing model (i.e., the actual pose information of the scraper conveyor after the previous push and before the current push, referred to as the actual pose information before the push). Predict the baseline of the scraper conveyor after the current push using the baseline prediction model (i.e., the predicted baseline of the scraper conveyor after the push before the push, referred to as the baseline of the scraper conveyor before the push). Then, based on the spatial difference obtained from the spatial difference feedback model, correct the predicted baseline of the scraper conveyor after the current push. Obtain the initial push control quantity by calculating the difference between the corrected scraper conveyor baseline and the pose information of the current scraper conveyor push section.

[0095] In this embodiment, the baseline of the scraper conveyor after the current push is predicted by the baseline prediction model, and its pose information is recorded as follows. The pose information of the current scraper conveyor push section is obtained from the working face information processing model and recorded as follows: Using Relations Solve for the attitude error, where This provides the yaw angle information for the scraper conveyor. This refers to the roll angle information of the scraper conveyor. For the pitch angle information of the scraper conveyor, Given the rotation matrix from A1 to A2; use the CalculateDeviation component in Unity3D software to calculate the positional error, where... The position information of the scraper conveyor is obtained; based on the calculated attitude error and position error, and the spatial difference obtained in step 402 of the spatial difference feedback model, a more accurate baseline of the scraper conveyor after pushing is obtained by calculation and analysis. Finally, the difference between the corrected scraper conveyor baseline and the current position and attitude information of the scraper conveyor pushing section is calculated to obtain a more accurate initial push control quantity.

[0096] In practice, the pseudocode for calculating the coordinate difference of X using the CalculateDeviation component is as follows:

[0097]

[0098]

[0099] In addition to the methods described above, the method of the present invention can also use the quaternion method to obtain the initial calculation control quantity.

[0100] Step 502, Floor Segmentation: Based on the initial calculated pusher control quantity and the coal seam floor update model based on physical behavior predicted in Step 204, obtain the floor data corresponding to the pusher section of the scraper conveyor, and the pusher trajectory information of the middle trough of each section of the scraper conveyor pusher; establish the floor function Z = F(x,y,z) corresponding to the pusher trajectory of the middle trough of each section of the scraper conveyor pusher based on the floor data, and find the extreme points of the floor function, letting... The process involves solving for points that may become extreme points and determining whether they are indeed extreme points. Ultimately, the extreme points of the coal seam floor corresponding to the pushing trajectory of the middle trough in each section of the scraper conveyor's pushing section are determined. Based on these extreme points, the coal seam floor in the pushing direction of the scraper conveyor's pushing section is segmented. Simultaneously, based on the corresponding extreme points, an optimized discretization method is used to segment the pushing trajectory of the middle trough in each section of the scraper conveyor's pushing section. This optimized discretization method is a known existing method that involves linking independent variables with target variables for discretization. In this invention, the target variable is the coordinate change ΔX of the pushing control quantity in the pushing direction, and the independent variable is the X-coordinate value of the coal seam floor.

[0101] Step 503: Simulate and solve the real-time pushing position of the middle trough of each section of the scraper conveyor in Unity3D software: Establish a coordinate system based on the pushing mechanism between the scraper conveyor and the hydraulic support, with the pushing direction of the scraper conveyor as the X-axis direction, such as... Figure 5 As shown, two points located on the side of the scraper conveyor, namely the pushing ear hole 6, are selected on the middle trough 5 of each section of the scraper conveyor as key points 1 and 2. The pose information of the two key points 1 and 2 is obtained, thereby obtaining a vector information corresponding to the middle trough. Based on the floating connection mechanism model in the pushing mechanism, the coordinates of the contact point 4 between the connecting head pin 3 of the floating connection mechanism and the pushing ear hole in the X-axis direction are calculated. Based on the above vector information, the position information of the contact point is calculated. Finally, based on the equipment posture information obtained by the coal seam floor update model based on physical behavior and the working face information processing model predicted in step 204, the real-time pushing position of the middle trough of each section of the scraper conveyor is calculated.

[0102] Step 504: Simulate the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software: Calculate the number of trough sections in the middle of the S-shaped curved section using the existing calculation method for the length of the curved section, and determine the number of sections in the front and rear curved sections (i.e., separate the front and rear curved sections). Establish the parent-child relationship of the middle troughs in Unity3D software and set a limit - i.e., the maximum curvature of the middle trough of each section. Then calculate the pushing amount required for the middle trough of each section of the corresponding shape of the scraper conveyor, and execute the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software.

[0103] In this embodiment, the process of establishing the parent-child relationship of the central groove is as follows: Five key point pins are added to the central groove, namely the lower right, upper right, lower left, upper left, and central. For the central groove of the first half of the curved section, the lower right pin is taken as the parent object, its coordinates are the coordinates of the previous lower left pin, and it rotates one degree around the pin. For the central groove of the second half, the upper right pin is taken as the parent object, its coordinates are the coordinates of the previous upper left pin, and it rotates one degree around the pin. When the next central groove is pushed, each central groove successively becomes the position of the previous central groove. This part of the technical content has been disclosed in the prior art.

[0104] Step 505, Control Quantity Screening Step: Based on the pushing simulation performed in Unity3D software in the above steps, output the segment information of each section of the scraper conveyor's middle trough that cannot be pushed smoothly during the entire pushing process; and determine whether there is a segment of the middle trough in each section of the scraper conveyor's pushing section that cannot be pushed smoothly at each instant during the pushing process. If any segment cannot be pushed smoothly, the corresponding middle trough pushing control quantity information is eliminated until the segment information of each section of the scraper conveyor's pushing section that can be pushed smoothly, as well as the corresponding middle trough pushing control quantity information, are obtained.

[0105] Step 506: Construct the cost function and find the theoretical minimum: Based on the segmented selection of base plate feature points in step 502, the cost function is constructed as follows:

[0106]

[0107] Where x(t) is the current position of the scraper conveyor; x t `arget` represents the position of the starting point of the corresponding segment on the base plate (i.e., the dividing point between two segments); `t` represents the time taken for the scraper conveyor to push the conveyor across the corresponding segment on the base plate; `u(t)` represents the distance taken for the scraper conveyor to push the conveyor across the corresponding segment on the base plate. 2 α represents the force exerted by the hydraulic cylinder during the pushing process; α and β represent the proportions of positional cost and behavioral cost. This method focuses on the control variables during the process, so α = 0.7 and β = 0.3.

[0108] Based on the initial calculation of the push-pull control quantity in step 501, the theoretical minimum value is calculated according to the cost function listed in equation (1);

[0109] Step 507: Selection of Optimal Push Control Quantity: Based on the push control quantity of the central trough obtained in Step 505 and the cost function constructed in Step 506, calculate the cost function under different pushing distances and pushing forces, and compare it with the theoretical minimum value to determine whether the minimum value can be reached (i.e., less than or equal to the minimum value). If the minimum value is reached, directly derive the corresponding push control quantity of the central trough as the optimal push control quantity. If the minimum value is not reached, re-segment according to the extreme point in Step 502. Based on the extreme point, adjust the segmentation point position and translate it towards the push start direction to complete the re-segmentation. Based on the re-segmentation of the coal seam floor, adjust the pushing distance and pushing force of the scraper conveyor push section, and repeat the above simulation and judgment process in Unity3D software. Through continuous adjustment and optimization, the cost function reaches the minimum value.

[0110] Step 508: Formulate the optimal push-pull strategy: Formulate the optimal push-pull strategy based on the optimal push-pull control quantity obtained in step 507; and transmit the optimal push-pull strategy to the feedback control model of the execution space.

Claims

1. A method for straightening a scraper conveyor based on the concept of rolling time-domain control, characterized in that: It includes the following spaces: the execution space containing the feedback control model, the derivation space containing the working face information processing model, and the prediction space containing the coupled base plate update model, the baseline prediction model, the spatial difference feedback model, and the control quantity optimization model. The feedback control model in the execution space receives the optimal pushing strategy of the scraper conveyor finally determined by the prediction space. Based on the optimal pushing strategy, the electro-hydraulic control system controls the subsequent pushing operation of the scraper conveyor to achieve the purpose of straightening the scraper conveyor. At the same time, the real-time information of the working face is fed back to the working face information processing model in the simulation space. The working face information processing model in the simulation space calculates and inverts the real-time working face information fed back by the execution space feedback control model after it runs, and sends the simulation information to the prediction space; the real-time working face information includes the position and posture information of the three fully mechanized mining machines, the cutting information of the coal mining machine, and the information of the coal seam floor. The prediction space receives the simulation information from the deduction space to perform simulation prediction and finally determine the optimal pushing strategy of the scraper conveyor; wherein, the coupled floor update model constructs a virtual known coal seam floor based on the simulation information, establishes the coupling relationship between the fully mechanized mining equipment and the coal seam floor, and finally predicts and updates the coal seam floor model of the area to be mined. The baseline prediction model is based on the aforementioned coupled base plate update model. It simulates the pushing process of the scraper conveyor and predicts the baseline of the scraper conveyor after pushing, providing a calculation basis for the subsequent spatial difference feedback model and control quantity optimization model. The spatial difference feedback model provides correction data for the equipment posture adjustment of the subsequent control quantity optimization model; the spatial difference refers to the spatial difference between the actual posture information of the scraper conveyor after the current push and the baseline of the scraper conveyor after the previous push predicted by the baseline prediction model. The control quantity optimization model obtains the optimal push-pull strategy based on the coupled base plate update model, baseline prediction model, and spatial difference feedback model, and provides it to the feedback control model of the execution space.

2. The method for straightening a scraper conveyor based on the concept of rolling time-domain control according to claim 1, characterized in that: The specific process by which the working face information processing model acquires the position and pose information of the three fully mechanized mining machines is as follows: Step 101: Hydraulic Support Group Pose Information Acquisition Steps: Install a 3D LiDAR at the middle of the coal mining machine body and scan the hydraulic support group during the coal mining machine cutting process; take the pin shaft connecting the base and the push cylinder on the hydraulic support, and the vertex position of the connection between the top beam and the side guard plate as feature parts, acquire the point cloud data of the two feature parts, and perform filtering, segmentation, and registration operations; calculate the relatively accurate point cloud data to finally obtain the pose information of the hydraulic support group; Step 102, Scraper Conveyor Position Information Acquisition Steps: Acquire data from the strapdown inertial navigation system on the coal mining machine; eliminate the accumulated error of the strapdown inertial navigation system using the extended Kalman filter method to obtain accurate coal mining machine position information; based on the positional relationship between the scraper conveyor and the coal mining machine, invert the scraper conveyor trajectory using the coal mining machine position information; select key points on the middle trough of each section of the scraper conveyor and use the key points to calculate the scraper conveyor position information; there are no special requirements for the selection of the key points, generally the center point of the middle trough of each section of the scraper conveyor is selected; Step 103, Data Processing Steps: Eliminate outliers in the pose information obtained in the above steps; quantify and store the information data to facilitate subsequent information deduction and update operations.

3. The method for straightening a scraper conveyor based on the concept of rolling time-domain control according to claim 1, characterized in that: The specific construction process of the coupled base plate update model is as follows: Step 201: Construct a virtual coal seam floor: Based on the known coal seam floor information obtained from the working face information processing model, construct a virtual coal seam floor using the Mesh component in Unity3D software; Step 202: Construct a coupling relationship model between the virtual coal seam floor and the equipment model: Import the equipment model into Unity3D software and set relevant rigid body components for the equipment model; adjust the position parameters of the equipment model so that the equipment model and the virtual coal seam floor can fit together fully, thus completing the construction of the coupling relationship model between the virtual coal seam floor and the equipment model; the equipment includes a coal mining machine, a hydraulic support group, and a scraper conveyor; Step 203: Predict the corrected model of the coal seam floor after the scraper conveyor pushes the conveyor: Based on the coal cutting information of the coal mining machine and the coal falling information in the past cutting process of the coal mining machine obtained by the working face information processing model, the deep LSTM neural network method is applied to predict the corrected model of the coal seam floor after the scraper conveyor pushes the conveyor. Step 204: Predict the coal seam floor update model based on physical behavior: Based on the coupling relationship model in Step 202 and the coal seam floor correction model in Step 203, set the scraper conveyor pushing amount, run Unity3D software to simulate the scraper conveyor pushing operation, apply the SURF algorithm to extract the feature point information of the coal seam floor corresponding to the pushing section after the scraper conveyor pushes, reconstruct the coal seam floor corresponding to the pushing section after the scraper conveyor pushes, and obtain the coal seam floor reconstruction model; based on the coal seam floor reconstruction model, analyze the impact of the scraper conveyor pushing behavior on the pits and loose coal piles of the coal seam floor. The failure scenarios were analyzed, and the stress conditions were assessed. Based on the Mohr-Coulomb criterion, the failure criteria for the coal seam floor after the scraper conveyor pushed the conveyor were derived. Based on the failure criteria, the coal seam floor reconstruction model was numerically simulated using FLAC3D software. The maximum depth and location of the damage caused to the coal seam floor by the support pressure along the pushing direction and lateral side of the scraper conveyor under the stress model were calculated. Finally, based on the coal seam floor reconstruction model and the results of the maximum damage depth and location, a physical behavior-based coal seam floor update model was constructed using the Mesh component in Unity3D software.

4. The method for straightening a scraper conveyor based on the concept of rolling time-domain control according to claim 1, characterized in that: The specific construction process of the baseline prediction model is as follows: Step 301: Represent the position information of the key points of the middle trough of each section of the scraper conveyor based on the first-level coordinate system; there are no special requirements for the selection of the key points of the middle trough, and the center point of the middle trough of each section of the scraper conveyor is generally selected. Step 302: Compare the coordinate information of each section of the middle trough in the pushing direction of the scraper conveyor, determine the last section of the middle trough, and record its sequence number i; Step 303: Based on the updated coal seam floor model, establish a parallel system in Unity3D software to simulate the pushing process of the scraper conveyor and obtain the position information of the last remaining section of the middle trough after it has been pushed to its full stroke. Step 304: Select n position points evenly throughout the entire pushing stroke of the last section of the middle groove, label the n position points from 1 to n in the direction from the start of pushing to the end of pushing, and obtain the position information of each position point; Step 305: Using the position line corresponding to position point n as the end point of the pushing stroke, simulate the pushing process of the remaining middle troughs of the scraper conveyor in the aforementioned parallel system, and determine whether the key points of the other middle troughs can all reach the position line corresponding to position point n. If they can all reach it, then the position line is the predicted baseline of the scraper conveyor after pushing. If they cannot all reach it, select the position line corresponding to the previous position point as the end point of the pushing stroke, and repeat the above simulation and judgment process until the key points of the remaining middle troughs of the scraper conveyor can all reach the end point of the pushing stroke.

5. The method for straightening a scraper conveyor based on the concept of rolling time-domain control according to claim 1, characterized in that: The spatial difference feedback model consists of three parts: feedforward space, feedback space, and correction mechanism. The specific construction process is as follows: Step 401, Feedforward Space: Based on the aforementioned baseline prediction model, predict the process of the scraper conveyor pushing the baseline in the past, extract the prediction rules in the prediction process, calculate the possible situations, and take corresponding measures to eliminate possible deviation problems in advance. Step 402, Feedback Space: Based on the baseline of the scraper conveyor after the push and the actual pose information of the scraper conveyor after the push based on the baseline prediction model mentioned above, the spatial difference is obtained by calculation; Step 403, Correction Mechanism: Combining the data information obtained from the feedforward space and the feedback space, the correlation coefficient is applied to the feedforward space prediction process to obtain the corrected prediction result; Continuously monitor feedback data and make adjustments based on real-time feedback information; continuously iterate and optimize the correction mechanism based on feedback results from actual operations.

6. The method for straightening a scraper conveyor based on the concept of rolling time-domain control according to claim 1, characterized in that: The specific construction process of the control quantity optimization model is as follows: Step 501: Obtain the initial push control quantity: Obtain the position and orientation information of the current scraper conveyor push section from the working face information processing model, predict the baseline of the scraper conveyor after the current push from the baseline prediction model, and then correct the predicted scraper conveyor baseline after the current push from the spatial difference feedback model based on the spatial difference. By calculating the difference between the corrected scraper conveyor baseline and the position and orientation information of the current scraper conveyor push section, the initial push control quantity is obtained. Step 502, Floor Segmentation: Based on the initial calculated pusher control quantity and the coal seam floor update model based on physical behavior predicted in Step 204, obtain the floor data corresponding to the pusher section of the scraper conveyor, and the pusher trajectory information of the middle trough of each section of the scraper conveyor pusher section; establish the floor function Z=F(x,y,z) corresponding to the pusher trajectory of the middle trough of each section of the scraper conveyor pusher section based on the floor data, and find the extreme points of the floor function, letting... The process involves solving for points that may become extreme points and determining whether they are indeed extreme points. Ultimately, the extreme points of the coal seam floor corresponding to the pushing trajectory of the middle trough in each section of the scraper conveyor are determined. Based on the determined extreme points, the coal seam floor in the pushing direction of the scraper conveyor is segmented. At the same time, the pushing trajectory of the middle trough in each section of the scraper conveyor is segmented using an optimized discretization method based on the corresponding extreme points. Step 503: Simulate and solve the real-time pushing position of each section of the scraper conveyor in Unity3D software: Establish a coordinate system based on the pushing mechanism between the scraper conveyor and the hydraulic support. Take the pushing direction of the scraper conveyor as the X-axis direction. Select two points on the side of the scraper conveyor, namely the pushing ear hole (6), on the middle section (5) of each section of the scraper conveyor as key point 1 and key point 2. Obtain the pose information of key point 1 and key point 2 to obtain a vector information corresponding to the middle section. Based on the floating connection mechanism model in the pushing mechanism, calculate the coordinates of the contact point (4) between the connecting head pin (3) of the floating connection mechanism and the pushing ear hole in the X-axis direction. Calculate the position information of the contact point based on the above vector information. Finally, calculate the real-time pushing position of each section of the scraper conveyor based on the equipment posture information obtained from the coal seam floor update model based on physical behavior and the working face information processing model predicted in step 204. Step 504: Simulate the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software: Calculate the number of trough sections in the middle of the S-shaped curved section using the existing calculation method for the length of the curved section, and determine the number of sections in each of the front and rear curved sections. Establish the parent-child relationship of the middle trough in Unity3D software and set a limit - that is, the maximum curvature of the middle trough of each section. Then calculate the pushing amount required for the middle trough of each section of the corresponding shape of the scraper conveyor, and execute the pushing operation of the S-shaped curved section of the scraper conveyor in Unity3D software. Step 505, Control Quantity Screening Step: Based on the pushing simulation performed in Unity3D software in the above steps, output the segment information of each section of the scraper conveyor's middle trough that cannot be pushed smoothly during the entire pushing process; and determine whether there is a segment of the middle trough of each section of the scraper conveyor's pushing section that cannot be pushed smoothly at each instant during the pushing process. If any section cannot be pushed smoothly, the corresponding middle trough pushing control quantity information is eliminated until the segment information of each section of the scraper conveyor's pushing section that can be pushed smoothly, as well as the corresponding middle trough pushing control quantity information, are obtained. Step 506: Construct the cost function and find the theoretical minimum: Based on the segmented selection of base plate feature points in step 502, the cost function is constructed as follows: Equation (1) in, x ( t () indicates the current position of the scraper conveyor; x t arget This indicates the position of the starting point of the corresponding segment on the base plate; t The time it takes for the scraper conveyor to push the slide through the corresponding section of the bottom plate; u ( t ) 2 The force exerted by the pushing cylinder during the pushing process; α、β To determine the proportion of location costs to behavioral costs, this method focuses on considering the control variables during the process. α =0.7, β =0.3; Based on the initial calculation of the push-pull control quantity in step 501, the theoretical minimum value is calculated according to the cost function listed in equation (1); Step 507: Selection of Optimal Push Control Quantity: Based on the push control quantity of the central trough obtained in Step 505 and the cost function constructed in Step 506, calculate the cost function under different pushing distances and pushing forces, and compare it with the theoretical minimum value to determine whether the minimum value can be reached. If the minimum value is reached, directly derive the corresponding push control quantity of the central trough as the optimal push control quantity. If the minimum value is not reached, re-segment according to the extreme point in Step 502. Based on the extreme point, adjust the segmentation point position and translate it towards the push start direction to complete the re-segmentation. Based on the re-segmentation of the coal seam floor, adjust the pushing distance and pushing force of the scraper conveyor push section. Repeat the above simulation and judgment process in Unity3D software. Through continuous adjustment and optimization, until the cost function reaches the minimum value. Step 508: Formulate the optimal push-pull strategy: Formulate the optimal push-pull strategy based on the optimal push-pull control quantity obtained in step 507; and transmit the optimal push-pull strategy to the feedback control model of the execution space.

Images