A method for collecting coordinates of a memory cutting track of a boom-type tunneling machine

Abstract

The application discloses a kind of cantilever type tunneling machine memory cutting track coordinate collection methods, it is related to intelligent equipment technical field, including step one: parameter setting, step two: sampling, step three: initialization, step four: steering identification, step five: track record, step six: track use, through "turning point collection" algorithm, accurately solve the data redundancy and inefficiency problem caused by timing collection, the core logic is "variable timing trigger is event trigger", no longer mechanically according to time grab coordinates, but through intelligent identification track geometric inflection point, only record "non-storing key point", solve the technical problem that the linear segment is stored in all power-off storage area by timing collection mode with a large number of repeated, meaningless intermediate point, cause storage resource waste, concentrate operation resource at the inflection point where truly need to change direction, greatly improve the dynamic response speed and execution efficiency of automatic cutting.

Description

Technical Field

[0001] This invention relates to the field of intelligent equipment technology, specifically a method for acquiring the coordinates of a cantilever tunneling machine's memory cutting trajectory. Background Technology

[0002] According to the evaluation method for the normalized operation of intelligent underground coal mines: during normal tunneling, for Class I mines, the tunneling machine memory cut-off rate shall not be less than 50%; for Class II mines, the tunneling machine memory cut-off rate shall not be less than 40%; and for Class III mines, the tunneling memory cut-off rate shall not be less than 30%.

[0003] Currently, most cantilever tunneling machines acquire their cutting trajectory coordinates using a "timed acquisition" method. This involves acquiring the cutting head coordinates (horizontal X, vertical Z) in real-time at unit time intervals throughout the driver's demonstration, and storing the data in the main controller's power-off storage area.

[0004] Taking a 5-minute demonstration with a 1-second acquisition cycle as an example, a total of 300 coordinate arrays need to be stored. Considering the speed of the demonstration and pauses at key locations, the arrays are filled with a large number of repetitive values. Given the existing intelligent control mode (control using a programmable controller and trajectory display using host computer software, both generally communicating via MODBUS_TCP protocol over Ethernet), the programmable controller needs a particularly large power-off storage area. At the same time, data transmission, trajectory plotting, and motion control for each sampling point will significantly reduce the time efficiency of automatic cropping. Summary of the Invention

[0005] The purpose of this invention is to provide a method for acquiring the memory cutting trajectory coordinates of a cantilever tunneling machine, in order to solve the problems proposed in the prior art, such as the pauses at speed and key positions, the large number of repetitive values ​​in the array, the need for a particularly large power-off storage area of ​​the programmable controller, and the significant reduction in the time efficiency of automatic cutting due to the data transmission, trajectory drawing, and motion control of each sampling point.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for acquiring the coordinates of a cantilever tunneling machine's memory cutting trajectory, comprising the following steps: Step 1: Parameter Settings Setting (1): The weight ratio W of the coordinate sampling points, W0>W1>W2>W3>W4>W5>W6>W7>W8>W9 and W0+W1+W2+W3+W4+W5+W6+W7+W8+W9=100%; Setting (2): Line judgment values ​​L1 and L2; Settings (3): Slant line judgment value k, slant line error range koff; Setting (4): Before the driver issues the formal "demonstration command", add a 3-5 second "no-load / load test switch" segment; The system collects the vibration acceleration of the cutting head and the pressure fluctuation of the hydraulic cylinder in real time. Based on the fluctuation amplitude, it automatically generates a basic threshold: L2=(peak value of fluctuation noise×2)+1cm, L1=2×L2, koff=dynamically calculated based on the standard deviation of the angle jitter when the cutter moves along an oblique path; it completely solves the shortcomings of "relying on human experience" and makes the threshold strongly bound to the current geological hardness and machine vibration characteristics.

[0007] Step 2: Sampling Create a sampling array [0..9] (data format: structure, containing X and Z coordinate values); The X and Z coordinates of the cutting head are collected once every 0.1 seconds. The values ​​of the array are shifted to the right, and the latest collected coordinate values ​​are stored in the array [0]. Current coordinates , .

[0008] Step 3: Initialization: First, place the cutting head at the target position when the automatic cutting is completed; the driver issues a demonstration command, collects the current X and Z coordinates of the cutting head, and assigns them to all addresses of the trajectory array [0..49] respectively; clear the count of steps and clear the turning status (0_meaningless, 1_horizontal, 2_vertical, 3_diagonal).

[0009] Step 4: Turning Recognition (1) Identification rules: Take n=0~49, the coordinates of the turning point Xn and Zn of the trajectory planning; calculate △X=|Xr-Xn| and △Z=|Zr-Zn|; identify when △X<L2; Step 5: Track Recording Status description: 0_meaningless, 1_horizontal, 2_vertical, 3_diagonal line; Recording strategy: (1) When the state changes to 1, 2 or 3 (rising edge), count the number of steps n→n+1, and store the current coordinates X and Z values ​​into the trajectory coordinate arrays X[n+1] and Z[n+1] (refreshed at all times); (2) When the state changes to 0, the number of steps remains unchanged, and access to the trajectory coordinate array is paused; Step Six: Track Usage: The recorded trajectory coordinates are stored in the programmable controller's power-off memory area, which serves the following two purposes: (1) The trajectory coordinate array is uploaded via Ethernet communication using the MODBUS_TCP protocol for display on the host computer; the number of data transmitted per cycle (format: WORD) is ≤100. (2) Used to automatically cut off the target operating point of PID control.

[0010] Preferably, in step one, L2 is the automatic tunnel cutting error of 10cm, and L1 is twice L2.

[0011] Preferably, during the sampling process in step two, the data validity watchdog module is activated: if the rate of change of Xr and Zr calculated by three consecutive samplings exceeds the physical limit (such as the maximum moving speed of the cutting head × sampling period), the sensor is determined to be disturbed, the previous valid value is forcibly maintained and an alarm is triggered, which greatly enhances the ability to resist strong vibrations and electromagnetic interference in the well and prevents the false triggering of turning point recording.

[0012] Preferably, in step three, after clearing the turning state, a state machine injection test is performed: a set of simulated △X and △Z signals are manually injected into the state machine to verify whether the state machine can jump normally from 0→1, 0→2, 0→3; If the transition is abnormal or times out (>100ms), immediately display "State machine initialization failed" and prevent automatic demonstration from starting; Preferably, in step four, a micro-motion holding area is set: In Table 1, when the state is 1 (horizontal) or 2 (vertical), if the changes in ΔX and ΔZ fall within a small hysteresis loop near L2 (such as 0.8L2-1.2L2), the original state remains unchanged, avoiding frequent switching back and forth at the threshold boundary, effectively avoiding the recording of "mistaken turning points" caused by single sampling jitter, and making the trajectory coordinates purer.

[0013] Preferably, during step four, if the state machine remains in any state for more than the preset longest straight-line time (e.g., 15 seconds) and ΔX and ΔZ are both less than L2, the state is automatically forced to return to zero to avoid deadlock: this solves the risk of state machine deadlock and improves system robustness.

[0014] Preferably, during the fifth step, a slope mutation check is performed: before storing a new turning point, the slope between the point and the previous turning point is calculated; if the slope deviates from the average of the last three slopes by more than 3×koff, the system determines it as an "abnormal jump point", refuses to store the point, and keeps the state unchanged (while reporting a "suspected trajectory distortion" prompt). Preferably, during step five, collinear point merging is performed: if three consecutively recorded turning points are approximately collinear (the perpendicular distance between the middle point and the line connecting the beginning and end is <1cm), the middle point is automatically removed, and only the beginning and end are retained. While ensuring that "only turning points are stored," minor jitter points from manual demonstrations are further removed, making the stored trajectories smoother and more refined.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. When this application is used, the "turning point acquisition" algorithm accurately solves the problems of data redundancy and low efficiency caused by timed acquisition. The core logic is to "change timed triggering to event triggering". Instead of mechanically capturing coordinates according to time, it intelligently identifies the geometric inflection points of the trajectory and only records the key points that "must be saved".

[0016] 2. This application, through the turning recognition logic in step four (determining the relationship between △X, △Z and thresholds L1, L2), automatically filters out "meaningless" intermediate points. Coordinates are only recorded when the state changes from a rising edge (i.e., from 0 to 1, 2, or 3). This solves the technical problem of wasting storage resources caused by storing a large number of repetitive and meaningless intermediate points on straight segments in the power-off storage area during timed acquisition. For the same 2-minute rectangular cross-sectional trajectory, timed acquisition requires storing 120 sets of data, while this solution only requires 18 sets (including inflection points only). The data volume dropped by more than 85%, greatly relieving the pressure on the power-off storage area and reducing the number of data groups from 120 to 18. The amount of data to be transmitted in a single session (format: WORD) is far lower than the "≤100 data points / cycle" limit specified in step six. This solves the technical problem that excessive data volume directly leads to Ethernet transmission needing to be split into multiple cycles, occupying bus bandwidth. Timed acquisition requires 3 Ethernet transmission cycles to complete, while turn point acquisition only requires 1 cycle, improving communication efficiency by 3 times and avoiding frequent occupation of the MODBUS_TCP bus.

[0017] 3. In this application, during automatic truncation, the PID controller needs to run each stored target point sequentially. If 120 points are stored, the PID controller needs to run 120 times, and the trajectory coordinates are used for the "target running points of automatic truncation PID control". Since only 18 key turning points are stored, the PID controller only needs to process these 18 core inflection points. Straight segments can maintain direction through slope interpolation or by directly maintaining direction. The number of PID runs is reduced to 18, and the controller does not need to perform frequent fine-tuning calculations. The computing resources are concentrated on the inflection points that actually need to change direction. This solves the technical problem that a large number of points within straight segments do not contribute to the trajectory direction and only increase the computing burden, greatly improving the dynamic response speed and execution efficiency of automatic truncation. Attached Figure Description

[0018] Figure 1 The cross-sectional shape of the tunnel being excavated; Figure 2 When △X < L2, the curve that maintains its original state cannot be identified. Figure 1 ; Figure 3 When △X < L2, the curve that maintains its original state cannot be identified. Figure 2 ; Figure 4 When △X < L2, the curve is identified as the curve in the state of transitioning from vertical to vertical. Figure 5 When L2≤△X<L1, the curve diagram that maintains the original state cannot be identified; Figure 6 When L2≤△X<L1, the curve is identified as having a low slope and maintaining its original state. Figure 7 The curve diagram identified as a diagonal line turning into a diagonal line when L2≤△X<L1; Figure 8 When L2≤△X<L1, the curve is identified as a curve in a state where the slope is high and then turns into a sloping line. Figure 9 When △X≥L1, it is identified as a curve in the state of transitioning from horizontal to horizontal. Figure 10 When △X≥L1, the curve is identified as a curve in the state of transitioning from a slope to a slope. Figure 1 ; Figure 11 When △X≥L1, the curve is identified as a curve in the state of transitioning from a slope to a slope. Figure 2 ; Figure 12 This is a schematic diagram of the cutting path; Figure 13 This is a schematic diagram of the tool path for edge sweeping. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: Figures 1-13 As shown, this invention provides a method for acquiring the coordinates of a cantilever tunneling machine's memory cutting trajectory: Step 1: Parameter Settings Setting (1): The weight ratio W of the coordinate sampling points, W0>W1>W2>W3>W4>W5>W6>W7>W8>W9 and W0+W1+W2+W3+W4+W5+W6+W7+W8+W9=100%; Setting (2): Straight line judgment values ​​L1 and L2, where L2 is the automatic roadway cutting error of 10cm and L1 is twice L2; Settings (3): Slant line judgment value k, slant line error range koff; Setting (4): Before the driver issues the formal "demonstration command", add a 3-5 second "no-load / load test switch" segment; The system collects the vibration acceleration of the cutting head and the pressure fluctuation of the hydraulic cylinder in real time. Based on the fluctuation amplitude, it automatically generates a basic threshold: L2=(peak value of fluctuation noise×2)+1cm, L1=2×L2, koff=dynamically calculated based on the standard deviation of the angle jitter when the cutter moves along an oblique path; it completely solves the shortcomings of "relying on human experience" and makes the threshold strongly bound to the current geological hardness and machine vibration characteristics.

[0021] Step 2: Sampling Create a sampling array [0..9] (data format: structure, containing X and Z coordinate values); The X and Z coordinates of the cutting head are collected once every 0.1 seconds. The values ​​of the array are shifted to the right, and the latest collected coordinate values ​​are stored in the array [0]. Current coordinates , ; During the sampling process, the data validity watchdog module is activated: if the rate of change of Xr and Zr calculated by three consecutive samplings exceeds the physical limit (such as the maximum moving speed of the cutting head × sampling period), the sensor is determined to be disturbed, the previous valid value is forcibly maintained and an alarm is triggered, which greatly enhances the ability to resist strong vibrations and electromagnetic interference in the well and prevents false triggering of turning point recording.

[0022] Step 3: Initialization: First, place the cutting head at the target position when automatic cutting is completed; the driver issues a demonstration command, collects the current X and Z coordinates of the cutting head, and assigns them to all addresses in the trajectory array [0..49] respectively; clear the count of steps and the steering status (0_meaningless, 1_horizontal, 2_vertical, 3_diagonal). After clearing the transition state, perform a state machine injection test: manually inject a set of simulated △X and △Z signals into the state machine to verify whether the state machine can jump normally from 0→1, 0→2, 0→3; if the jump is abnormal or timeout (>100ms), immediately prompt "state machine initialization failed" and prohibit the start of automatic demonstration.

[0023] Step 4: Turning Recognition (1) Identification rules: take n=0~49, the coordinates of the turning point Xn and Zn of the trajectory planning; calculate △X=|Xr-Xn| and △Z=|Zr-Zn|; identify when △X<L2; set a micro-motion holding area during the identification process: in Table 1, when the state is 1 (horizontal) or 2 (vertical), if the change of △X and △Z falls within a small hysteresis loop near L2 (such as 0.8L2-1.2L2), keep the original state unchanged, avoid frequent switching at the threshold boundary; effectively avoid the "mistaken turning point" record caused by single sampling jitter, and make the trajectory coordinates purer. If the state machine stays in any state for more than the preset longest straight line segment time (such as 15 seconds) and △X and △Z are both less than L2, then the state is automatically forced to return to zero, avoid deadlock, solve the deadlock risk of the state machine, and improve the robustness of the system.

[0024] The steering identification analysis table is as follows: Table 1: Steering Recognition in States 0, 1, and 2

[0025] The slope of each newly identified diagonal line is k0, meaning that k0 is not a fixed value; Table 2: Steering Recognition in State 3

[0026] Step 5: Track Recording Status description: 0_meaningless, 1_horizontal, 2_vertical, 3_diagonal line; Recording strategy: (1) When the state changes to 1, 2 or 3 (rising edge), count the number of steps n→n+1, and store the current coordinates X and Z values ​​into the trajectory coordinate arrays X[n+1] and Z[n+1] (refreshed at all times); (2) When the state changes to 0, the number of steps remains unchanged, and access to the trajectory coordinate array is paused; In the above process, collinear points are merged: if three consecutively recorded turning points are approximately collinear (the perpendicular distance between the middle point and the line connecting the beginning and end is less than 1cm), the middle point is automatically removed, and only the beginning and end are retained.

[0027] While ensuring that "only inflection points are stored", we further eliminate minor jitter points during human demonstrations to make the stored trajectory smoother and the quantity more refined. Simultaneously, slope mutation verification is performed: before storing a new turning point, the slope between the point and the previous turning point is calculated; if the slope deviates from the average of the three most recent slopes by more than 3×koff, the system determines it as an "abnormal jump point", refuses to store the point, and keeps the state unchanged (while reporting a "suspected trajectory distortion" prompt).

[0028] Step Six: Track Usage: The recorded trajectory coordinates are stored in the programmable controller's power-off memory area, which serves the following two purposes: (1) The trajectory coordinate array is uploaded via Ethernet communication using the MODBUS_TCP protocol for display on the host computer; the number of data transmitted per cycle (format: WORD) is ≤100. (2) Used to automatically cut off the target operating point of PID control.

[0029] In summary, this application uses the "turning point acquisition" algorithm for testing and comparison. Taking the acquisition of the memory cutting trajectory of a rectangular cross section of the same specification as an example, the core logic is to "change timed triggering to event triggering". Instead of mechanically capturing coordinates according to time, it intelligently identifies the geometric inflection points of the trajectory and only records the key points that "must be stored", which accurately solves the problems of data redundancy and low efficiency caused by timed acquisition.

[0030] Driver demonstration and explanation: The whole process takes about 2 minutes, with a total of 2 diagonal cutting segments, 7 vertical cutting segments, and 8 horizontal cutting segments, and a total of 18 turning points (including start and end points). The comparison results are shown in Table 3 below. Table 3: Comparison Table

[0031] This application uses turning point acquisition technology to collect and memorize the cutting trajectory, which greatly reduces the power failure storage area of ​​the stored data (120 sets > 18 sets), improves the data upload speed (3 Ethernet transmission cycles > 1 Ethernet transmission cycle), and improves the efficiency of automatic cutting control (120 PID running times > 8 PID running times).

[0032] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for acquiring coordinates of a cantilever tunneling machine's cutting trajectory, characterized in that: Includes the following steps; Step 1: Parameter Settings Setting (1): The weight ratio W of the coordinate sampling points, W0>W1>W2>W3>W4>W5>W6>W7>W8>W9 and W0+W1+W2+W3+W4+W5+W6+W7+W8+W9=100%; Setting (2): Line judgment values ​​L1 and L2; Settings (3): Slant line judgment value k, slant line error range koff; Setting (4): Before the driver issues the formal "demonstration command", add a 3-5 second "no-load / load test switch" segment; The system collects the vibration acceleration of the cutting head and the pressure fluctuation of the hydraulic cylinder in real time. Based on the fluctuation amplitude, it automatically generates the basic threshold: L2=(peak value of fluctuation noise×2)+1cm, L1=2×L2, koff=dynamically calculated based on the standard deviation of the angle jitter when the blade moves along the oblique line. Step 2: Sampling Create a sampling array [0..9] (data format: structure, containing X and Z coordinate values); The X and Z coordinates of the cutting head are collected once every 0.1 seconds. The values ​​of the array are shifted to the right, and the latest collected coordinate values ​​are stored in the array [0]. Current coordinates , .

2. Step 3: Initialization: First, place the cutting head at the target position when the automatic cutting is completed; the driver issues a demonstration command, collects the current X and Z coordinates of the cutting head, and assigns them to all addresses of the trajectory array [0..49] respectively; clear the count of steps and clear the turning status (0_meaningless, 1_horizontal, 2_vertical, 3_diagonal).

3. Step Four: Turn Recognition: (1) Identification rules: Take n=0~49, the coordinates of the turning point Xn and Zn of the trajectory planning; calculate △X=|Xr-Xn| and △Z=|Zr-Zn|; identify when △X<L2; Step 5: Track Recording Status description: 0_meaningless, 1_horizontal, 2_vertical, 3_diagonal line; Recording strategy: (1) When the state changes to 1, 2 or 3 (rising edge), count the number of steps n→n+1, and store the current coordinates X and Z values ​​into the trajectory coordinate arrays X[n+1] and Z[n+1] (refreshed at all times); (2) When the state changes to 0, the number of steps remains unchanged, and access to the trajectory coordinate array is paused; Step Six: Track Usage: The recorded trajectory coordinates are stored in the programmable controller's power-off memory area, which serves the following two purposes: (1) The trajectory coordinate array is uploaded via Ethernet communication using the MODBUS_TCP protocol for display on the host computer; the number of data transmitted per cycle (format: WORD) is ≤100. (2) Used to automatically cut off the target operating point of PID control.

4. The method for acquiring coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: In step one, L2 is the automatic tunnel cutting error of 10cm, and L1 is twice L2.

5. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: In step two, during the sampling process, the data validity watchdog module is activated: if the rate of change of Xr and Zr calculated by three consecutive samplings exceeds the physical limit (such as the maximum moving speed of the cutting head × sampling period), the sensor is determined to be disturbed, the previous valid value is forcibly maintained and an alarm is triggered, thereby enhancing the ability to resist strong vibrations and electromagnetic interference downhole and preventing false triggering of turning point recording.

6. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: In step three, after clearing the turning state, a state machine injection test is performed: a set of simulated △X and △Z signals are manually injected into the state machine to verify whether the state machine can jump normally from 0→1, 0→2, 0→3.

7. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: In step four, a micro-motion holding area is set: In Table 1, when the state is 1 (horizontal) or 2 (vertical), if the changes in ΔX and ΔZ fall within a small hysteresis loop near L2 (such as 0.8L2-1.2L2), the original state remains unchanged to avoid frequent switching back and forth at the threshold boundary.

8. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: During step four, if the state machine remains in any state for more than the preset longest straight-line segment time (e.g., 15 seconds) and both ΔX and ΔZ are less than L2, the state will be automatically forced to return to zero to avoid deadlock.

9. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: During the fifth step, a slope mutation check is performed: before storing a new turning point, the slope between the current point and the previous turning point is calculated; if the slope deviates from the average of the last three slopes by more than 3×koff, the system determines it as an "abnormal jump point".

10. The method for acquiring the coordinates of a cantilever tunneling machine's cutting trajectory according to claim 1, characterized in that: During the fifth step, collinear points are merged: if three consecutively recorded turning points are approximately collinear (the perpendicular distance between the middle point and the line connecting the beginning and end is less than 1cm), the middle point is automatically removed, and only the beginning and end are retained.

Images