JP6949623B2 - Operation setting presentation device and program - Google Patents

Description

The present invention relates to an operation setting presentation device and a program.

Conventionally, when constructing a tunnel or the like by the soil pressure shield method (mud pressure shield method), the construction environment (soil quality, water pressure, etc.) of the site excavated by the shield excavator changes from moment to moment depending on the location of the site. .. Therefore, the operator manually operates the shield excavator while comparing the pre-planned excavation instruction with the measurement data from various measuring devices measured according to the construction environment and monitoring the measurement data. I do. For example, in the earth pressure type shield method, the operator monitors the earth pressure in the chamber at the time of excavation and discharges the soil generated by excavation so that the earth pressure in the chamber falls within a predetermined range. For example, the operator controls the amount of soil discharged in a predetermined time by manually changing the set value of the rotation speed of the screw conveyor, and operates so that the earth pressure in the chamber is within a predetermined range ( For example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2013-199772

However, in order to adjust the rotation speed of the screw conveyor so that the earth pressure in the chamber is within a predetermined range, it is necessary to balance the excavation speed and the earth discharge speed by the screw conveyor. The excavation speed is closely related to a large number of data such as cutter torque and total thrust. Therefore, it is necessary to operate the rotation speed of the screw conveyor while monitoring a large amount of data, and the operation is difficult unless the operator has a high degree of skill.
The number of skilled operators is decreasing, it is difficult to secure them, and it takes a lot of time to request highly skilled operators. In addition, even a skilled operator may make a mistake in the operation timing and set values, and if the shield excavator is not operated properly, the accuracy and safety of the excavated tunnel design will decrease. There are concerns.
Therefore, there is a demand for a method in which the set value of the operation in the shield excavator is appropriately presented.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide an operation setting presenting device and a program capable of showing accurate operation setting values in a shield excavator. ..

In order to solve the above-mentioned problems, in the operation setting presentation device of the embodiment of the present invention, the setting data regarding the operation setting value of the shield excavator in the excavation situation is added to the excavation status data regarding the excavation status of the shield excavator. It is equipped with a model creation unit that creates a setting prediction model that predicts the set value of the operation in the shield excavator by executing machine learning using the training data, and the excavation status data includes the construction environment of the shield excavator. Environmental data related to, and operation data related to the excavation operation of the shield excavator corresponding to the environmental data are included, the environmental data includes indicated soil pressure and control soil pressure, and the operation data includes shield excavation. It is characterized in that the measured value relating to the screw conveyor of the machine is included, and the setting data includes the setting value of the operation relating to the screw conveyor.

Further, in the operation setting presentation device according to the embodiment of the present invention, the model creation unit uses the data selected from the data to which the setting data is attached to the excavation status data as training data to perform machine learning. It is characterized by executing.

Further, in the operation setting presentation device according to the embodiment of the present invention, the model creation unit selects and selects the excavation status data in which the difference between the controlled earth pressure and the indicated earth pressure is equal to or less than a predetermined threshold value. It is characterized in that machine learning is executed using the learning data in which the setting data is attached to the excavation status data.

Further, the operation setting presentation device according to the embodiment of the present invention uses the excavation status data acquisition unit that acquires the excavation status data at predetermined time intervals and the setting prediction model created by the model creation unit. It is further provided with a prediction unit that predicts a set value of an operation in the shield excavator in the situation shown in the excavation status data, and an output unit that outputs a prediction result predicted by the prediction unit.

Further, the operation setting presentation device according to the embodiment of the present invention sets the setting data acquisition unit for acquiring the setting data at predetermined time intervals and the drilling status data acquired by the excavation status data acquisition unit. A storage unit that stores data associated with the setting data acquired by the data acquisition unit is further provided, and the prediction unit causes the model creation unit to execute machine learning using the data stored in the storage unit. This is characterized by updating the setting prediction model.

Further, the program of one embodiment of the present invention is a program for causing a computer to function as the above-mentioned operation setting presenting device.

As described above, according to the present invention, it is possible to show the set value of the operation with high accuracy in the shield excavator.

It is a figure explaining the shield excavator 20 of the object which the operation setting presenting apparatus 100 of embodiment creates a setting prediction model. It is a block diagram which shows the structural example of the operation setting presenting apparatus 100 of embodiment. It is a figure which shows the structural example of the learning data table in the learning data storage unit 107 of embodiment. It is a flowchart which shows the operation example of the operation setting presentation apparatus 100 of embodiment.

Hereinafter, the operation setting presenting device and the program of the embodiment will be described with reference to the drawings.

FIG. 1 is a diagram illustrating a shield excavator 20 for which the operation setting presentation device 100 (see FIG. 2) of the present embodiment creates a setting prediction model.
FIG. 1A shows a conceptual diagram of a shield excavator 20 for which the operation setting presentation device 100 of the present embodiment creates a setting prediction model. FIG. 1B is a conceptual diagram of a propulsion jack that propels the shield excavator 20.

As shown in FIG. 1 (a), the shield excavator 20 performs a primary lining S by assembling segments with an elector (not shown) at the rear portion of the cylindrical skin plate 2 in the direction of arrow D1. At the same time, it is a mechanism for excavating the ground 30. In the shield excavator 20, a mud chamber 7 is provided at the rear of the annular and face plate type cutter 10 provided with the cutter bit 23 in the direction of the arrow D1. A plurality of soil pressure gauges 70 are installed in the mud soil chamber 7. The soil pressure gauge 70 measures the mud pressure in the mud-making chamber 7. The mud-making soil material 9 is injected into the mud-making soil chamber 7 from the mud-making soil material injection pipe 8, and the excavated earth and sand is converted into mud by mixing with a kneading blade (not shown). In the shield excavator 20, the earth is stabilized and excavated by balancing the earth pressure of the mud-making chamber 7 with the earth pressure and the water pressure on the excavation surface (face) of the earth 30. The excavated residual soil accumulated in the mud-making chamber 7 is introduced into the screw conveyor 60 and discharged to the outside of the excavated tunnel via the conveyors 62 and 63. The gantry M supports each of the screw conveyor 60 and the conveyors 62 and 63.
Further, the shield excavator 20 improves the fluidity of the excavated soil by adding bubbles made of a special foaming material to the excavated soil instead of injecting the mud soil material 9, and the mud soil chamber 7 (chamber). In some cases, excavation is carried out by adjusting the earth pressure inside.

Here, in the screw conveyor 60, the rotation speed of the screw is controlled by hydraulic control. When the rotation speed of the screw of the screw conveyor 60 (hereinafter, simply referred to as the rotation speed of the screw) increases, the mud soil in the mud soil chamber 7 is discharged, so that the mud pressure decreases. The mud pressure rises due to the stagnation of the mud discharge. That is, the mud pressure is controlled by performing an operation of setting the rotation speed of the screw.

As shown in FIG. 1 (b), the propulsion jack is arranged between the skin plate 2 and the segment so as to surround the inner circumference of the skin plate 2. When the propulsion jack is propelled (extended) by the flood control operation, the surface of the skin plate 2 is pushed and the shield excavator 20 propels.

Here, the force point for propelling the surface of the skin plate 2 is set depending on which position of the propulsion jack is propelled, and the propulsion direction of the shield excavator 20 is determined. Further, the propulsion speed of the shield excavator 20 is determined by the speed at which the propulsion jack is propelled. That is, the propulsion direction of the shield excavator 20 is controlled by performing an operation of setting which propulsion jack is selected. Further, the propulsion speed of the shield excavator 20 is controlled by performing an operation of setting the propulsion speed of the propulsion jack. That is, in the shield excavator 20, the earth pressure is controlled by balancing the excavation speed of the shield excavator 20 and the soil discharge speed accompanying the rotation of the screw, and the earth pressure is controlled to stabilize the ground. And excavate.

In the present embodiment, excavation status data regarding the excavation status at the time of excavation of the shield excavator 20 is acquired. The excavation status data includes environmental data regarding the construction environment of the shield excavator and operation data regarding the excavation operation of the shield excavator corresponding to the environmental data.
Environmental data are, for example, indicated earth pressure and controlled earth pressure. Here, the indicated earth pressure is the earth pressure of the ground 30 working in the direction of the arrow A, and is determined in consideration of the influence of the subsidence of the ground on the earth pressure derived based on the depth and the geology. The controlled earth pressure is a representative value of the earth pressure inside the mud-making chamber 7. For example, the controlled earth pressure is an average value of the earth pressure measured every predetermined time (for example, every second) of the earth pressure gauge 70 installed near the center of the mud making chamber 7.
The operation data includes, for example, the measured value of the rotation speed of the screw, the oil pressure of the screw conveyor 60, the opening degree of the gate (exit) of the screw conveyor 60, the stroke of the propulsion jack, the torque of the cutter 10, the propulsion speed of the shield excavator 20 and the like. The total propulsion force of the propulsion jack, the bubble injection rate, etc. These environmental data and operation data are measured at predetermined time intervals (for example, every second).

Further, in the present embodiment, setting data relating to the operation of the shield excavator 20 is acquired. The setting data is, for example, a set value of the rotation speed of the screw, a set value of selection of the propulsion jack, and a set value of the propulsion speed of the propulsion jack. These setting data are acquired at the same time as the predetermined time when the operation data is measured.

FIG. 2 is a block diagram showing a configuration example of the operation setting presentation device 100 of the embodiment.
The operation setting presentation device 100 includes, for example, an excavation status data input unit 101, a setting data input unit 102, a prediction result output unit 103, a processing unit 104, a model creation unit 105, a setting prediction unit 106, and learning data. A storage unit 107 is provided.

The excavation status data input unit 101 is, for example, at the timing when a timekeeping signal indicating the passage of a predetermined time (for example, 1 second) from the timer (not shown) of the operation setting presenting device 100 is supplied, the excavation status data input unit 101 described above. To get. For example, the measurement result measured by the soil pressure gauge 70 installed near the center of the mud making chamber 7 is input to the excavation status data input unit 101. Further, in the excavation status data input unit 101, for example, the detection values detected by the screw conveyor 60 and various sensors (not shown) installed on the propulsion jack are input. The excavation status data input unit 101 stores the acquired excavation status data in association with the acquired time (time stamp) in the learning data table in the learning data storage unit 107.

The setting data input unit 102 acquires the above-mentioned setting data at the timing when the excavation status data is acquired by the excavation status data input unit 101, for example. In the setting data input unit 102, setting values of various operations input to a touch panel or the like on the operation panel (not shown) of the shield excavator 20 are input. The setting data input unit 102 stores the acquired setting data in the learning data table in the learning data storage unit 107 in association with the acquired time.

The prediction result output unit 103 outputs the content of the operation predicted by the setting prediction unit 106. The prediction result output unit 103 is, for example, a display device, and displays the content of the operation predicted by the setting prediction unit 106. Further, the prediction result output unit 103 is, for example, a speaker, and outputs the content of the operation predicted by the setting prediction unit 106 by voice, an alarm sound, or the like.

The processing unit 104 generates learning data based on the excavation status data acquired by the excavation status data input unit 101 and the setting data acquired by the setting data input unit 102. The processing unit 104 extracts, for example, a feature amount in a plurality of time-series excavation status data. The feature amount is a characteristic component included in the excavation status data, for example, the difference between the indicated earth pressure and the controlled earth pressure, the value obtained by moving average the excavation status data for a short time (for example, 10 seconds), or a long time (for example). For example, it can be obtained by calculating a moving average value (60 seconds). The processing unit 104 stores the extracted feature amount in the learning data table in the learning data storage unit 107 in association with the corresponding time.

In addition, the processing unit 104 determines whether or not to use the excavation status data and the setting data acquired at each timing as learning data. Judgment as to whether or not to use as training data is performed based on, for example, the degree of deviation between the indicated earth pressure and the controlled earth pressure.
When the difference between the indicated earth pressure and the controlled earth pressure at a certain time (hereinafter referred to as the earth pressure difference) is equal to or more than a predetermined threshold value, the processing unit 104 does not use the excavation status data and the setting data at that time as learning data. Is determined. Further, when the earth pressure difference at a certain time is less than a predetermined threshold value, the processing unit 104 determines that the excavation status data and the setting data at that time are used as learning data. The processing unit 104 stores the result of the determination performed on the learning data table in the learning data storage unit 107 in association with the corresponding time.
The processing unit 104 may make a determination based not only on the earth pressure difference but also on the frequency and variation of operations. Further, the processing unit 104 may make a determination based on the result of cluster analysis of the excavation status data and the setting data.

The model creation unit 105 performs setting prediction for predicting the set value of the operation in the shield excavator 20 by executing machine learning using the data stored in the training data table in the training data storage unit 107 as learning data. Create a model. Here, as the machine learning technique for creating the setting prediction model, any of commonly used techniques such as decision tree learning, neural network, genetic programming, and support vector machine may be used.
Further, the model creation unit 105 executes machine learning using the excavation status data determined to be used as the training data by the processing unit 104 and the setting data corresponding to the excavation status data as learning data. As a result, the model creation unit 105 can learn, for example, the excavation status data and the setting data when the earth pressure is controlled more appropriately and accurately.

The setting prediction unit 106 predicts the set value in the operation of the shield excavator 20 in the situation shown in the excavation status data by using the setting prediction model created by the model creation unit 105. The setting prediction unit 106 uses the data output from the setting prediction model as the data indicating the predicted operation when the excavation status data acquired at a certain time is input to the setting prediction model.

The learning data storage unit 107 stores a learning data table for causing the model creation unit 105 to create a setting prediction model.
FIG. 3 is a diagram showing a configuration example of a learning data table in the learning data storage unit 107 of the embodiment. In the example of FIG. 3, the learning data table has each item of time stamp, excavation status data, excavation status data short-time moving average, excavation status data long-term moving average, earth pressure difference, setting data, and classification. The time stamp stores the time when the excavation status data and the setting data were acquired. The excavation status data acquired at the time indicated by the time stamp is stored in the excavation status data. In addition, each of a plurality of excavation status data (excavation status data # 1, # 2, ..., # N) is stored in the excavation status data. Where N is an arbitrary natural number.

The short-time moving average value of each excavation status data for a short time (for example, 10 seconds) is stored in the excavation status data short-time moving average. The excavation status data long-time moving average stores the long-term (for example, 60 seconds) moving average value of each excavation status data. The difference between the indicated earth pressure and the controlled earth pressure is stored in the earth pressure difference. The excavation status data acquired at the time indicated by the time stamp is stored in the setting data. Each of a plurality of setting data (excavation status data # 1, ..., # M) is stored in the setting data. Here, M is an arbitrary natural number. A number indicating the classification performed by the processing unit 104 is stored in the classification, and for example, the number 0 indicates that the classification is less frequently used as learning data than the number 3.

FIG. 4 is a flowchart showing an operation example of the operation setting presentation device 100 of the embodiment.
First, the operation setting presentation device 100 acquires learning data (step S1). Here, the learning data acquired by the operation setting presentation device 100 may be excavation status data and setting data in the shield excavator 20 to be predicted by the operation setting presentation device 100 in operation, or other data. It may be standard excavation status data and setting data at the time of excavation in a shield excavator.

Next, the operation setting presentation device 100 creates a setting prediction model by executing machine learning using the acquired learning data (step S2).
Next, the operation setting presenting device 100 acquires the excavation status data and the setting data of the shield excavator 20 for which the set value in the operation is predicted (step S3). The acquired excavation status data and setting data are stored in the learning data storage unit 107. Further, the operation setting presentation device 100 stores the feature amount extracted based on the acquired excavation status data and the information indicating the classification as the learning data in the learning data storage unit 107, thereby storing the learning data storage unit 107. The training data is stored in the training data table of (step S4).

Next, the operation setting presentation device 100 determines whether or not to update the setting prediction model (step S5). The operation setting presentation device 100 determines that the setting prediction model is updated when an amount of learning data necessary for creating a model is accumulated in the learning data table in the learning data storage unit 107.
When the operation setting presentation device 100 determines that the setting prediction model is to be updated, the operation setting presentation device 100 updates the setting prediction model (step S6). The operation setting presentation device 100 causes the model creation unit 105 to execute machine learning using the learning data stored in the learning data table in the learning data storage unit 107, thereby creating a new setting prediction model. Then, the operation setting presentation device 100 updates the setting prediction model by replacing the setting prediction model with a new setting prediction model. Here, the newly created setting prediction model is a model created by using the excavation status data and the setting data acquired along with the excavation of the shield excavator 20. Therefore, the operation setting presenting device 100 can present a more accurate operation setting value by making a highly accurate prediction according to the excavation environment and the characteristics of the shield excavator 20 to be predicted. Become.

Then, the operation setting presenting device 100 predicts the set value in the operation of the shield excavator 20 by inputting the acquired excavation status data into the setting prediction model (step S7). The operation setting presentation device 100 outputs data indicating the predicted operation setting value (step S8).

As described above, the operation setting presenting device 100 of the embodiment performs machine learning with respect to the operation of the shield excavator 20 by using the learning data in which the setting data in the excavation situation is attached to the excavation status data. A model creation unit 105 for creating a setting prediction model for predicting a set value in the operation of the shield excavator 20 is provided.
Thereby, the operation setting presenting device 100 of the embodiment can show the set value of the operation with high accuracy in the shield excavator 20. This is because the model creation unit 105 can create a model showing the relationship between the excavation status and the set value of the operation. By presenting the set value output by inputting the excavation status to the model created by the model creation unit 105, it is possible to show the set value of the operation with high accuracy in the shield excavator 20.

Further, in the operation setting presenting device 100 of the embodiment, the excavation status data includes environmental data regarding the construction environment of the shield excavator 20 and operation data regarding the excavation operation of the shield excavator 20 corresponding to the environmental data. As a result, the operation setting presentation device 100 of the embodiment can create a setting prediction model based on the environment of the excavation site and the characteristics of the shield excavator 20, and shields based on the setting data output from the model. It is possible to show the set value of the operation with high accuracy in the excavator 20.

Further, in the operation setting presenting device 100 of the embodiment, the environmental data includes the indicated earth pressure at the time of excavation and the indicated earth pressure, and the operation data includes the measured value regarding the screw conveyor 60 of the shield excavator 20. The setting data includes the setting value of the operation related to the screw conveyor 60. As a result, the operation setting presentation device 100 of the embodiment can create a setting prediction model based on the respective relationships between the highly relevant indicated earth pressure, the indicated earth pressure, and the set value of the screw conveyor 60. Based on the setting data output from the model, it is possible to show the setting value of the operation with high accuracy in the shield excavator 20.

Further, in the operation setting presentation device 100 of the embodiment, the model creation unit 105 executes machine learning using the data determined to be used for the learning data based on the difference between the indicated earth pressure and the controlled earth pressure. As a result, the operation setting presentation device 100 of the embodiment can create a setting prediction model using more appropriate data, and based on the setting data output from the model, the operation setting presentation device 100 of the shield excavator 20 can perform an accurate operation. The set value can be shown.
Further, in the operation setting presentation device 100 of the embodiment, the model creation unit 105 selects the excavation status data in which the earth pressure difference is equal to or less than a predetermined threshold value, and uses the learning data in which the setting data is added to the selected excavation status data. Use to perform machine learning. As a result, the operation setting presentation device 100 of the embodiment creates a setting prediction model using more appropriate data by using the data when the earth pressure is appropriately controlled to be closer to the target value as training data. It is possible to show the set value of the operation with high accuracy in the shield excavator 20 based on the setting data output from the model.

Further, the operation setting presenting device 100 of the embodiment is shown in the excavation status data by using the excavation status data input unit 101 that acquires the excavation status data at predetermined time intervals and the setting prediction model created by the model creation unit. It further includes a setting prediction unit 106 that predicts a set value in the operation of the shield excavator 20 in the above situation, and a prediction result output unit 103 that outputs the prediction result predicted by the setting prediction unit 106. As a result, the operation setting presenting device 100 of the embodiment can predict the set value in the operation by the setting prediction unit 106 based on the excavation status of the shield excavator 20 acquired by the excavation status data input unit 101. The predicted operation can be output by the prediction result output unit 103. Therefore, the operation setting presenting device 100 of the embodiment can present the recommended value of the operation set by the operator, and can function as, for example, an operation support device for executing the operation guidance. Further, automatic operation is possible by converting the recommended operation value into a setting signal on the operation panel of the shield excavator 20 and outputting it to the operation panel.

Further, the operation setting presenting device 100 of the embodiment is acquired by the setting data input unit 102 that acquires the setting data at predetermined time intervals and the setting data input unit 102 that acquires the excavation status data acquired by the excavation status data input unit 101. A learning data storage unit 107 that stores data associated with the set setting data is further provided, and the setting prediction unit 106 causes the model creation unit 105 to execute machine learning using the data stored in the learning data storage unit 107. By updating the setting prediction model. As a result, the operation setting presenting device 100 of the embodiment is set to correspond to the excavation status data accumulated as the excavation progresses and the excavation status data of the shield excavator 20 for which the set value in the operation is predicted. The setting prediction model can be updated based on the data, and accurate prediction can be performed. Therefore, it is possible to show the set value of the operation with high accuracy in the shield excavator 20.

All or part of the processing performed by the operation setting presenting device 100 in the above-described embodiment may be realized by a computer. In that case, the program for realizing this function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read by the computer system and executed. The term "computer system" as used herein includes hardware such as an OS and peripheral devices. Further, the "computer-readable recording medium" refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, or a CD-ROM, or a storage device such as a hard disk built in a computer system. Further, a "computer-readable recording medium" is a communication line for transmitting a program via a network such as the Internet or a communication line such as a telephone line, and dynamically holds the program for a short period of time. It may also include a program that holds a program for a certain period of time, such as a volatile memory inside a computer system that serves as a server or a client in that case. Further, the above program may be for realizing a part of the above-mentioned functions, and may be further realized for realizing the above-mentioned functions in combination with a program already recorded in the computer system. It may be realized by using a programmable logic device such as FPGA.

Although the embodiments of the present invention have been described in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs and the like within a range that does not deviate from the gist of the present invention.

100 ... Operation setting presentation device, 101 ... Excavation status data input unit, 102 ... Setting data input unit, 103 ... Prediction result output unit, 104 ... Processing unit, 105 ... Model creation unit, 106 ... Setting prediction unit, 107 ... Learning data Memory.

Claims (6)

Setting the operation of the shield excavator by executing machine learning using the learning data in which the setting data of the operation of the shield excavator in the drilling situation is added to the excavation status data of the shield excavator. It has a model creation unit that creates a setting prediction model that predicts the value .
The excavation status data includes environmental data relating to the construction environment of the shield excavator and operation data relating to the excavation operation of the shield excavator corresponding to the environmental data.
The environmental data includes indicated earth pressure and controlled earth pressure.
The operation data includes measurements on the screw conveyor of the shield excavator.
The setting data includes setting values of operations related to the screw conveyor.
An operation setting presentation device characterized by this. The operation according to claim 1 , wherein the model creation unit executes machine learning by using data selected from the data to which the setting data is attached to the excavation status data as training data. Setting presentation device. The model creation unit selects the excavation status data in which the difference between the control earth pressure and the indicated earth pressure is equal to or less than a predetermined threshold value, and the training data in which the setting data is added to the selected excavation status data. The operation setting presenting device according to claim 1 or 2 , wherein machine learning is performed using the above. An excavation status data acquisition unit that acquires the excavation status data at predetermined time intervals,
Using the setting prediction model created by the model creation unit, a prediction unit that predicts the set value of the operation of the shield excavator in the situation shown in the excavation status data, and
The operation setting presenting device according to any one of claims 1 to 3 , further comprising an output unit that outputs a prediction result predicted by the prediction unit. A setting data acquisition unit that acquires the setting data at each predetermined time, and a setting data acquisition unit.
A storage unit for storing data in which the excavation status data acquired by the excavation status data acquisition unit is associated with the setting data acquired by the setting data acquisition unit is further provided.
The operation setting presentation device according to claim 4 , wherein the prediction unit updates the setting prediction model by causing the model creation unit to execute machine learning using the data stored in the storage unit. .. A program for causing a computer to function as the operation setting presenting device according to any one of claims 1 to 5.

Images