JP2023093812A - Operation method of shield machine and shield machine - Google Patents

Abstract

To provide an operation method of a shield machine which enables excavation work by automatic operation as much as possible even when an N value along an excavation track cannot be continuously grasped.SOLUTION: The present invention comprises an N-value estimation step of estimating N-values along an excavation track based on intermittent N-value measurements and/or geological information, and a control value derivation step for deriving a recommended value for a predetermined operation factor of the shield machine required for automatic operation in an excavation range where the N value estimated in the N value estimation step is less than a predetermined threshold based on the soil pressure in front of the excavation direction. The control value derivation step predicts the soil pressure ahead in the excavation direction and derives a recommended value for the operation factor of the shield machine based on the predicted soil pressure value.SELECTED DRAWING: Figure 5

Description

The present invention relates to a method of operating a shield machine and a shield machine.

When constructing a tunnel or the like using the shield construction method, the construction environment (ground conditions such as soil quality and water pressure) at the site where the shield excavator excavates changes moment by moment depending on the location of the site. Therefore, it is necessary for the operator to operate the speed and direction of excavation by the shield excavator in response to changes in the construction environment.

However, there is a problem that the accuracy and safety of the design of the tunnel to be excavated varies depending on the skill level of the operator who operates the shield excavator.

In Patent Document 1, by modeling the operation of a shield excavator by a skilled operator based on the relationship with monitoring data of the corresponding construction environment, it is possible to achieve operational support and uniformity of operation. A shield excavator operation analysis system has been proposed.

The shield excavator operation analysis system relates to the operation of the shield excavator, and corresponds each of the monitoring item data for monitoring the excavation status of the shield excavator and the operation value of the skilled operator's operation corresponding to each of the monitoring item data. A shield excavator operation analysis system for analyzing a relationship, comprising: an operation data change detection unit for detecting a change in the operation value; A monitoring item data detection unit to be extracted, and obtaining a correlation between a change amount of the operation value and a change amount of each of the monitoring item data extracted at the operation value change timing of the operation value, thereby obtaining the monitoring item data. A data relevance analysis unit for extracting monitoring item data related to the operation value from each of the detection units.

Cutter torque, cutter speed, propulsion pressure, propulsion speed, propulsion speed instruction deviation value, controlled earth pressure, face earth pressure average value outside the instruction range dummy, agitator torque, screw speed, primary screw pressure, Secondary screw pressure, NO. 1 copy stroke, NO. 1 copy stroke indicated value difference, NO. 1 copy position, NO. 1 copy position instruction sheet deviation value, pitching, pitching instruction value difference, rolling, rolling instruction value difference, top/bottom center-bending angle, top/bottom center-bending angle difference, left/right center-bending angle, left/right center-bending angle difference, S/ M forward heading, S/M forward heading difference, S/M rear heading, S/M rear heading difference, planned route horizontal deviation (control point), planned route vertical deviation (control point), Direction (control point), direction indication value difference (control point), planned route direction (control point), pitch (control point), planned route pitch (control point), etc., are subject to monitoring. A model is created with monitoring item data as explanatory variables and operation data as explained variables.

In other words, by creating a model between the input explanatory variables and the output dependent variables using machine learning, and using this model, highly related dependent variables can be identified according to the combination of explanatory variables. make it available. By using this model, it is possible to clearly obtain which of the operation data, which is the explained variable, needs to be manipulated in response to changes in the monitoring item data, which is the explanatory variable.

Further, Patent Document 2 proposes a control device for realizing automatic excavation control for a boring excavator.
The control device includes: output parameter acquiring means for acquiring each value of one or more output parameters output from the excavator; a determination means for predicting and determining the stratum or hardness of the stratum; a first control mode for automatic excavation prepared in advance for each stratum type of the stratum; and a second control mode for automatic excavation prepared in advance for each depth of the stratum. and a control means for executing excavation control of the excavator by switching to the first control mode or the second control mode according to the stratum or hardness predicted and judged by the judgment means. there is

Japanese Patent Application Laid-Open No. 2018-021402 JP 2019-206906 A

However, when actually excavating using a shield machine, it is affected by the N value (standard penetration test value), which is the standard for the compactness and strength of the excavated soil. Proper excavation was not guaranteed. In particular, the inventors of the present invention have found that favorable results cannot be obtained in hard soil with a large N value.

On the other hand, since there are areas where boring surveys are difficult due to the influence of ground structures, etc., it is extremely difficult to continuously grasp the N value that fluctuates along the excavation course.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method of operating a shield machine and a shield machine that enable automatic excavation work as much as possible even when the N value along the excavation course cannot be continuously grasped in view of the above-described situation. It is in the point to do.

To achieve the above objects, a first feature of the method of operating a shield machine according to the present invention is to estimate the N-value along the excavation track based on intermittent N-value measurements and/or geological information. an N-value estimation step, and in an excavation range in which the N-value estimated in the N-value estimation step is less than a predetermined threshold value, an automatic and a control value derivation step of deriving recommended values of predetermined operating parameters of the shield machine necessary for operation.

To execute an N-value estimating step of estimating the N-value based on intermittent N-values obtained over the excavation course even when the continuous N-values cannot be actually measured over the excavation course. , a continuous N-value can be obtained over the excavation track. Then, in the control value derivation step, based on the estimated value of the N value, the soil pressure in the excavation direction forward is controlled in advance within a range in which the N value is less than a predetermined threshold and the soil is estimated to be soft. Appropriate automated driving is enabled by deriving recommended values for driving control factors so that they are maintained within the range of values.

In addition to the above-described first characteristic configuration, the second characteristic configuration is that, in an excavation range in which the N value estimated in the N value estimation step is less than the threshold value, the operation The point is that automatic operation is performed according to the recommended value of the operation factor, and manual operation is performed by the operator in the excavation range in which the N value exceeds the threshold.

In the excavation range in which the N value is less than a predetermined threshold value in the N value estimation step and is estimated to be soft soil, it is appropriate to perform automatic operation using the recommended value of the driving factor derived in the control value derivation step. In addition, the excavation work proceeds smoothly, the N value exceeds the threshold value, and the excavation range is estimated to be hard soil.

In addition to the above-described first or second characteristic configuration, the third characteristic configuration is that the N-value estimation step obtains the intermittent N-value actually measured values and/or geological information along the excavation course. The point is that the interpolated value is estimated as the N value over the excavation course.

As the N-value over the excavation track, it is preferable to adopt intermittent measured N-values and/or interpolated values obtained by interpolating geological information over the excavation track. As an interpolation algorithm, for example, linear interpolation, Lagrange interpolation, spline interpolation, etc. can be used.

The fourth characteristic configuration is, in addition to any one of the first to third characteristic configurations described above, the control value derivation step predicting the earth pressure ahead in the excavation direction, and based on the predicted earth pressure value is to derive the recommended value of the operation factor of the shield machine.

By predicting the soil pressure ahead of the excavation direction and deriving the recommended value of the appropriate shield machine operation factor corresponding to the predicted soil pressure value, good automatic operation becomes possible.

The fifth feature configuration is, in addition to the above-described fourth feature configuration, excavation using a learned model constructed by performing machine learning using a data set of actual values of at least soil pressure and predetermined operation factors. The point is to predict the earth pressure in the forward direction.

In order to predict the soil pressure ahead of the excavation direction, a trained model constructed by performing machine learning using at least the actual values of soil pressure and predetermined driving factors as a data set is used to predict the appropriate driving factors. is sought.

A shield machine according to the present invention is characterized in that it comprises a rotating body with an excavating blade provided on the front surface of an outer cylinder, the rotating body is driven by a motor, and excavation is performed while the rear end of the outer cylinder is pressed by a pressing mechanism. A shield machine comprising a control device for controlling the rotating body and the pressing mechanism by executing the shield machine operation method having any one of the first to fifth characteristic configurations described above. It is in.

As described above, according to the present invention, there is provided a method of operating a shield machine and a shield machine that enable excavation work to be automatically operated as much as possible even when the N value along the excavation course cannot be continuously grasped. It became possible.

Illustration of the internal structure of the shield machine (a) is an explanatory diagram of the N-value estimation step, and (b) is an explanatory diagram of the threshold for separating automatic operation and manual operation based on the estimated N-value. Explanatory diagram of the machine learning device Characteristic diagram of predicted values predicted by a machine learning device based on actual measurements Flowchart showing how to operate a shield machine

The method of operating the shielding machine and the shielding machine according to the invention are described below.

"Shield construction method" refers to a construction method in which a shield machine is pushed in with a shield jack installed at the rear to excavate the ground and construct a tunnel. , This is a construction method in which the segment receives the reaction force of the shield jack and further propels the shield machine.

As shown in FIG. 1, the shield machine 1 is provided with a cutter head 3 that is rotationally driven by a cutter drive motor 4 on the front part of the main body section 2 . A shield jack 5 is arranged at the rear part of the main body 2, and the rotationally driven cutter head 3 is pressed by the shield jack 5 receiving a reaction force from the segment 13, thereby excavating the ground.

The earth and sand excavated by the cutter head 3 are taken into the cutter chamber 6 through the earth and sand intake port formed in the cutter head 3 . The cutter chamber 6 has a sealed structure separated from the main body 1 by a partition wall 7, and the cutter chamber 6 is filled with earth and sand fluidized by the excavation additive injected from the excavation additive injection port 8, and the cutter is produced. A pressure is held which opposes the earth pressure of C. The cutter head 3 and the cutter chamber 6 constitute an excavating section.

The earth and sand taken into the cutter chamber 6 are uniformly mixed with the excavation additive in the cutter chamber 6 to be plastically fluidized, and the earth and sand are conveyed through the earth-discharging screw conveyor 9 whose tip protrudes into the chamber 6. It is conveyed and discharged through a discharge valve provided at the rear end of the screw conveyor 9 . Reference numeral 11 is a screw conveyor drive motor.

In addition, the shield machine 1 includes a folding jack 12 for direction correction and curved construction, a stabilizer for preventing rolling, a hydraulic unit as a hydraulic source for the jacks, a power supply unit as a power supply for the electric motor, and a control device. It has

Values of various sensors that convey the state of the shield machine are input to the control device, and by controlling the above-described cutter drive motor 4, screw conveyor drive motor 11, soil discharge valve 10, various jacks 5, 13, 14, etc. , the shield machine 1 is propelled along a predetermined planning line. Various sensors include a soil pressure gauge, an inclinometer, a pressure gauge provided in the hydraulic system, a stroke gauge for detecting the stroke of various jacks, an encoder for detecting the number of rotations of the motor, an ammeter for detecting the current of the motor, and the like.

The method of operating the shield machine executed by the control device will be described in detail below. The operation method of the shield machine 1 includes an N-value estimation step of estimating the N-value along the excavation course based on intermittent N-value measurements and/or geological information, and the N-value estimated in the N-value estimation step In the excavation range that is less than the predetermined threshold, the recommended value of the predetermined operation factor of the shield machine necessary for automatic operation is derived so that the soil pressure in the front of the excavation direction is maintained within the preset control value range. and a control value derivation step.

As the main operating factors, the thrust by the shield jack 5, the cutter current of the cutter drive motor 4, the gate opening of the discharge valve, and the rotation speed of the screw conveyor are adopted.

As shown in FIG. 2(a), when the N values at points A and B along the excavation route, which is the planned line, are found to be a and b, respectively, by a preliminary boring survey, etc., the control device , an N value estimation step of estimating an N value at an arbitrary point between points A and B is executed.

That is, the N-value estimating step is configured to estimate intermittent N-values actually measured and/or interpolated values obtained by interpolating the geological information along the excavation route as the N-values along the excavation route. Instead of actually measuring the N value over the entire excavation section, intermittent measured values of the N value are interpolated over the excavation course. As an interpolation algorithm, the linear interpolation method is adopted in the example of FIG. By doing so, the accuracy of interpolation can be improved.

The N value that forms the basis of the N value estimation is not limited to the value obtained by a preliminary boring survey, and may be, for example, an N value included in geological information surveyed for academic purposes in the past. Moreover, when boring survey is difficult, the N value near the excavation route can be adopted instead of the N value on the excavation route.

The control device uses the machine learning device to estimate the soil pressure forward in the excavation direction in a range where the estimated N value obtained in the N value estimation step is less than a predetermined threshold and the soil is estimated to be soft. Automatic operation is performed by predicting and deriving recommended values for operation factors of the shield machine so that the soil pressure value is maintained within a preset control value range. On the other hand, in a range where the estimated N value exceeds a predetermined threshold value and is estimated to be hard soil, excavation is performed manually by an operator.

As shown in FIG. 2(b), when the N value of the intermediate position is obtained by linear interpolation based on the N values a and b obtained by the boring survey of the points A and B, the point between the points A and B When the estimated value at C is the threshold c, automatic operation is performed between points A and C where the N value is less than the threshold c, and manual operation is performed between points C and B where the N value is greater than the threshold c. will be driving.

The control value deriving step executed by the control device predicts the earth pressure at a predetermined distance forward in the excavation direction, and repeats the process of deriving the recommended value of the operation factor of the shield machine based on the predicted earth pressure value. It is something to do.

By predicting the earth pressure ahead of the excavation direction, it is possible to derive the recommended value of the appropriate shield machine operation factor corresponding to the predicted earth pressure value.

When predicting the earth pressure value, for example, a machine learning device employing random forest can be used. Using the thrust force, cutter current, gate opening, and screw conveyor rotation speed as explanatory variables, the machine learning device estimates the earth pressure value, which is the objective variable.

As shown in Fig. 3(a), thrust force, cutter current, gate opening, and screw conveyor rotation speed are used as explanatory variables, soil pressure is used as an objective variable, and actual measured values are prepared as learning data (teacher data). Then, the machine learning device is tuned so that the objective variable corresponding to the explanatory variable is output based on the learning data.

Next, as shown in Fig. 3(b), the initial values of the driving factors, which are thrust, cutter current, gate opening, and screw conveyor rotation speed, are input into the tuned machine learning device, and the Estimate the soil pressure value.
If the estimated earth pressure value does not fall within the allowable range of the preset control values, the operation factor is sequentially updated from the initial value, and the process of estimating the earth pressure value by the machine learning device is performed. A recommended value for the operating factor is determined repeatedly until the estimated value falls within the preset control value.

In this way, the control device automatically controls the shield machine based on the recommended value of the operation factor that allows the estimated earth pressure value to fall within the allowable range of the preset control value.

FIG. 4 shows the characteristics comparing the estimated value and the measured value when the soil pressure value is estimated using the machine learning device tuned in this way. The coefficient of determination R2 at this time was 0.83.

It has been found that it is possible to maintain the earth pressure value within the allowable range by automatically operating the shield machine based on the recommended values of the operating parameters obtained using a machine learning device.

FIG. 5 shows the procedure of operation control executed by the control device described above.
Based on intermittently obtained N values, the control device estimates the N value over the excavation course by linear interpolation (S1), and if the estimated N value is less than a predetermined threshold value (S2, Y), automatically The operation mode is entered, the initial values of operation factors are set (S3), and the soil pressure value ahead in the excavation direction is predicted by the machine learning device (S4). If the predicted earth pressure value does not fall within the preset control range (S5, N), the operation factor is set to the initial value until the predicted earth pressure value falls within the preset control range. While sequentially updating from , the processing from step S4 to step S5 is repeated. If the predicted soil pressure value falls within a preset control range (S5, Y), the setting value is determined as the recommended value for the operation factor (S6), and automatic operation control is executed (S7). ).

Until the preset excavation distance is reached (S8, N), the automatic operation control from step S2 to step S7 is continued, and when the preset excavation distance is reached (S8, Y), the estimation of step S2 Return to the N value determination process. Note that the preset excavation distance is a value that is appropriately set, and is not limited to a specific value.

In step S2, if the estimated N value is equal to or greater than the predetermined threshold value, it is determined that automatic operation is difficult (S2, N), the manual operation control mode is entered (S9), and the estimated N value is the predetermined threshold value. Manual operation control is continued until it becomes less than (S2, Y).

In the above-described embodiment, as automatic driving control, a machine learning device that uses a random forest is used to obtain a recommended value for a driving manipulation factor. may For example, it is possible to adopt a learning algorithm such as a neural network that learns using teacher data.

In the above-described embodiment, as automatic operation control, a machine learning device is used to obtain the recommended value of the operation factor. A recommended value of the driving manipulation factor may be calculated by executing a feedback calculation, and automatic driving control may be executed based on the calculation result.

In the above-described embodiment, four types of driving force, cutter current, gate opening, and screw conveyor rotation speed are adopted as operation factors that contribute highly to the prediction of the soil pressure value using the machine learning device. explained. However, the operation factor for maintaining the soil pressure value at the control value is not limited to these four types, and it is also possible to adopt the current of the electric motor that drives the screw conveyor.

Needless to say, in order to propel the shield machine 1 along a predetermined planned line, it is necessary to separately control the bending jack 12, the stabilizer, etc. in addition to the control of the soil pressure value.

Each of the above-described embodiments is an example of the present invention, and the technical scope of the present invention is not limited by the description, and it is possible to appropriately change and design within the range in which the effects of the present invention are exhibited. Needless to say.

1: Shield machine 2: Main body 3: Cutter head 4: Cutter drive motor 5: Shield jack 6: Cutter chamber 7: Partition wall 8: Excavation additive inlet 9: Screw conveyor C: Cutter CNT: Control device

Claims (6)

an N-value estimation step of estimating the N-value along the excavation track based on intermittent N-value measurements and/or geological information;
In the excavation range where the N value estimated in the N value estimation step is less than a predetermined threshold, the shield machine necessary for automatic operation so that the soil pressure ahead in the excavation direction is maintained within a preset control value range. a control value derivation step of deriving a recommended value for a predetermined driving maneuver factor of
How to operate a shield machine, including; In an excavation range in which the N value estimated in the N value estimation step is less than the threshold value, automatic driving is performed according to the recommended value of the driving factor derived in the control value derivation step, and the N value exceeds the threshold value. 2. The method of operating a shield machine according to claim 1, wherein manual operation is performed by an operator in an excavation range exceeding the excavation range. 3. The N value estimation step according to claim 1 or 2, wherein a value obtained by interpolating the intermittent N value actually measured and/or the geological information along the excavation course is estimated as the N value along the excavation course. How to operate a shield machine. 4. The control value deriving step predicts the earth pressure ahead in the excavation direction, and derives the recommended value of the operation factor of the shield machine based on the predicted earth pressure value. A method of operating the shield machine described in . 5. The method of operating a shield machine according to claim 4, wherein the earth pressure ahead of the excavation direction is predicted by a learned model constructed by performing machine learning using at least the earth pressure and the actual values of the predetermined operation factor as a data set. A shield machine that includes a rotating body having an excavating blade on the front surface of an outer cylinder, and excavates while rotating the rotating body and pressing the rear end of the outer cylinder with a pressing mechanism,
A shield machine comprising a control device that controls the rotating body and the pressing mechanism by executing the shield machine operating method according to any one of claims 1 to 5.

Images