JP2022045659A - Information processing system, information processing program, and learning device - Google Patents

Abstract

To provide an information processing system capable of supporting a user's determination.SOLUTION: An information processing system in an embodiment includes an acquisition unit, a model execution unit, and a display control unit. The acquisition unit acquires geological information relevant to a scheduled construction site. The model execution unit generates an estimated current value by inputting the geological information relevant to the scheduled construction site into a trained model that has learned a current value representing the excavation resistance when excavating the construction end location as the correct answer data by using as the input data the geological information relevant to the construction endpoint. The display control unit controls a display unit to display the estimated current value.SELECTED DRAWING: Figure 1

Description

The present invention relates to an information processing device, an information processing program and a learning device.

In pile construction to install piles in the ground, it is important to understand the geological condition of the planned construction site. In particular, it is important to know the depth of the support layer that supports the weight of the structure.

When constructing piles at the planned construction site, test excavation will be conducted at the planned construction site prior to the actual construction. In the test excavation, by actually excavating a hole and collecting fluctuations in the current value of the excavator's auger motor and earth and sand adhering to the excavation head, it is possible to check whether the geological information investigated in advance matches the actual geology. confirm. It is desirable to carry out test excavation evenly throughout the planned construction site, but in many cases, test excavation is carried out only at one to several places due to cost and construction period.

As a method for evaluating the ground, for example, in Patent Document 1, an input layer having an input element such as a current value and an excavation speed at each depth acquired in a test excavation and an N value acquired at a predetermined depth interval are output. A ground evaluation system that estimates the N value of a predetermined depth (hereinafter referred to as the estimated N value) for the input elements acquired during actual pile construction using a machine-learned N value calculation model based on layered teacher data is disclosed. Has been done.

Japanese Unexamined Patent Publication No. 2019-157346

In Patent Document 1, since N values measured at predetermined depth intervals, generally 1 m intervals, are used as teacher data, the measured values that can be obtained by excavating one hole are about several to several tens. .. Therefore, since the number of data acquired by the test excavation is small, there is a problem that the accuracy of the generated N value calculation model is low. Therefore, there is a problem that even if the estimated N value is calculated based on the current value and the excavation speed that can be obtained at the time of actual pile construction, the result reflecting the actual geology cannot be obtained.
Therefore, an object of the present invention is to estimate an assumed current value for a planned construction site based on a current value that can be acquired in large quantities as teacher data.

(1) The information processing apparatus according to one aspect of the present invention includes an acquisition unit, a model execution unit, and a display control unit. The acquisition department acquires geological information regarding the planned construction site. The model execution unit uses geological information about the construction end location as input data, and learns the current value indicating the excavation resistance when the construction end location is excavated as correct answer data. By inputting the geological information, an assumed current value is generated. The display control unit causes the display device to display the assumed current value.
According to the above configuration (1), since a large amount of current value, which is teacher data, can be acquired, it is possible to accurately generate a trained model in which the current value when excavating the construction end location is trained as correct answer data.
Since the trained model is highly accurate, the assumed current value for the planned construction site can be calculated with high accuracy.
Further, since the accuracy of the assumed current value is high, the user can accurately evaluate the ground while observing the estimated current value displayed by the display control unit.

(2) In some embodiments, in (1) above,
The correct answer data is data on a construction end place having a geology whose similarity with the geology of the planned construction place is equal to or more than a threshold value.
According to the above configuration (2), by using a trained model learned with data having similar geology, it is possible to obtain an assumed current value suitable for the geology of the planned construction site.

(3) In some embodiments, in the above configuration (1) or the above configuration (2).
The correct answer data is data relating to a construction end location within which the distance from the planned construction location is within a threshold value.
According to the above configuration (3), by using the trained model learned from the data near the planned construction site, it is possible to obtain an assumed current value suitable for the geology of the planned construction site.

(4) In some embodiments, in any one of the above configurations (1) to (3).
The display control unit displays the assumed current value and the current value obtained during excavation at the planned construction site in parallel.
According to the above configuration (4), the user's judgment can be supported by displaying the assumed current value and the current value during excavation in parallel so that the data can be seen at a glance. Specifically, it is possible to visually determine whether or not the excavation has reached the support layer.

(5) In some embodiments, in any one of the above configurations (1) to (3).
The display control unit superimposes and displays the assumed current value and the current value obtained during excavation at the planned construction site.
According to the above configuration (5), the user's judgment can be supported by superimposing and displaying the assumed current value and the current value during excavation so that the data can be understood at a glance. Specifically, it is possible to visually determine whether or not the excavation has reached the support layer.

(6) In the information processing program according to one aspect of the present invention, when a computer is used as an acquisition means for acquiring geological information regarding a planned construction site and geographic information regarding a construction end location is used as input data, and the construction end location is excavated. The model execution means for generating the assumed current value by inputting the geological information about the planned construction site to the trained model trained with the current value indicating the excavation resistance of the above as the correct answer data, and the assumed current value. It functions as a display control means to be displayed on the display device.
According to the above configuration (6), it is possible to support the judgment of the user.

(7) The learning device according to one aspect of the present invention has a storage unit for storing construction data in which geological information regarding the construction end location is associated with a current value indicating excavation resistance when the construction end location is excavated. , A learning unit that trains a model using learning data that uses the geological information about the construction end location as input data and the current value when excavating the construction end location as correct data, and generates a trained model. And.
According to the above configuration (7), it is possible to support the judgment of the user.

(8) In some embodiments, in the above configuration (7),
Further, it is provided with an acquisition unit for acquiring the geological information regarding the planned construction location, and a selection unit for selecting construction data related to the construction end location similar to the planned construction location as the learning data from the storage unit. The learning unit trains a model using the learning data selected by the selection unit, and generates the trained model.
According to the above configuration (8), a trained model can be generated according to the planned construction location, and a suitable assumed current value can be output.

(9) In some embodiments, in the above configuration (8),
The selection unit selects the construction data of the construction end place having the geology whose similarity with the geology of the planned construction place is equal to or more than the threshold value.
According to the above configuration (9), a trained model having a similar geology can be generated, and a suitable assumed current value can be output.

(10) In some embodiments, in the above configuration (8) or configuration (9).
The selection unit selects the construction data of the construction end place where the distance from the planned construction place is within the threshold value.
According to the above configuration (10), a trained model having similar locations can be generated, and a suitable assumed current value can be output.

According to the present invention, it is possible to accurately calculate the assumed current value for the planned construction site.

It is a block diagram which shows the information processing apparatus which concerns on this embodiment. It is a conceptual diagram at the time of model learning of the information processing apparatus which concerns on this embodiment. It is a conceptual diagram at the time of using the model of the information processing apparatus which concerns on this embodiment. It is a flowchart which shows the operation of the information processing apparatus which concerns on this embodiment. It is a flowchart which shows another example of the operation of the information processing apparatus which concerns on this embodiment. It is a figure which shows the display example of the output data of a trained model. It is a figure which shows another example of the display of the output data of a trained model. It is a flowchart which shows another example of the operation of the information processing apparatus which concerns on this embodiment. It is a conceptual diagram of the reproduction of the model learning of the information processing apparatus which concerns on this embodiment.

Hereinafter, the information processing apparatus, the information processing program, and the learning apparatus according to the embodiment of the present invention will be described in detail with reference to the drawings. In the following embodiments, the same operation is performed for the portions with the same number, and the description thereof will be omitted.

The information processing apparatus according to this embodiment will be described with reference to the block diagram of FIG.
The information processing apparatus 1 according to the present embodiment includes a processing circuit 11, a storage unit 12, and a communication interface 13, each of which is connected via a bus. The information processing device 1 according to the present embodiment may be mounted on a server, a PC, a tablet-type information terminal, a dedicated computer, or the like as an administrator terminal, or may be mounted on a management device mounted on a pile driving machine. Alternatively, the information processing apparatus 1 may be configured as a single device. When the information processing apparatus 1 is configured as a single device, it may include, for example, an input unit such as a keyboard, a mouse, and a touch panel, and a display unit such as a display.

The processing circuit 11 is realized by any one of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof. The processing circuit 11 includes an acquisition unit 111, a learning unit 112, a model execution unit 113, a display control unit 114, and a selection unit 115.

The acquisition unit 111 acquires position information and geological information regarding the planned construction site when the learned model described later is used. In the present embodiment, the "planned construction site" is assumed to be a location where piles are planned to be installed in the ground, but the present invention is not limited to this, and it is necessary to refer to actual measurement data when embedding an object in the ground such as an anchor. It can also be applied to work. The location information may be an address, or may be latitude and longitude coordinate information. The geological information includes at least geological data such as soil classification (for example, columnar chart) indicating the state of the geology along the depth direction and ground strength information (for example, N value, N value) indicating the degree of hardness of the ground in the depth direction. Compression strength, etc.) and included. The geological information may also include the date and time of the geological survey, the depth information determined to be the support layer, the groundwater level information, and the like.

The learning unit 112 learns the network model by using the position information and the geological information regarding the construction end place that has been completed as input data and the current value indicating the excavation resistance when excavating the construction end place as the correct answer data. Generate a trained model. The set of input data and correct answer data is called learning data. In the present embodiment, the "construction end location" is assumed to be a location having known geological information and current value on which the pile was constructed, but it is not limited to the construction of piles and may be a location having data such as geological information and current value. Just do it.
Further, in the present embodiment, the case where the integrated current value is used will be described, but the present invention is not limited to this. The integrated current value is an integrated value obtained by integrating the current value of the excavator's auger motor at predetermined depth intervals, and is calculated at predetermined depth intervals along the depth direction.

The model execution unit 113 outputs the assumed integrated current value by inputting the position information and the geological information regarding the planned construction location to the trained model. The assumed integrated current value is an assumed value of the integrated current value along the depth direction.

The display control unit 114 controls the display data including the assumed integrated current value to be displayed on a display device (not shown, for example, a screen via a display or a projector) via, for example, a communication interface 13 described later. The display control unit 114 may display the display data on the display device via a display interface (not shown) different from the communication interface 13.

The selection unit 115 selects training data according to a given condition. The method of selecting the learning data will be described later with reference to the flowcharts of FIGS. 4 and 5.

The storage unit 12 is composed of an HDD (Hard Disk Drive), SSD (Solid State Drive), etc., and stores construction data in which position information and geological information regarding the construction end location are associated with excavation information when the location is excavated. do.
Further, the storage unit 12 stores the position information and the geological information regarding the planned construction site in association with each other. In addition, training data, trained models, etc. are stored.
It is assumed that the excavation information regarding the construction end location includes at least time-series data of the current value and integrated current value of the motor, but data such as excavation date and time, excavator model information, excavation speed, and water supply amount are included. It may be included.

The communication interface 13 is an interface for transmitting / receiving data between the information processing device 1 and the external device, which conforms to a predetermined communication standard. Communication standards include mobile networks such as 4G and 5G, wireless LAN (Local Area Network) such as Wi-Fi (registered trademark), wireless networks such as Bluetooth (registered trademark) and NFC (Near Field Communication), or wired LAN. , USB (Universal Serial Bus) and other wired networks are assumed.

Next, the time of model learning of the information processing apparatus according to the present embodiment will be described with reference to the conceptual diagram of FIG. The storage unit 12, the learning unit 112, and the selection unit 115 are collectively referred to as a learning device.

The learning unit 112 uses learning data 21 as input data 21 for position information and geological information regarding the construction end location for which the geological survey has been completed, and correct data 22 for the integrated current value actually measured when the construction end location is actually constructed. The network model 20 is trained to generate the trained model 23. By training in this way, the trained model 23 is trained so that the integrated current value for the geological information can be output.

As a learning method, machine learning may be performed using a general method. For example, the network model may use a neural network, linear regression, random forest, or the like. Further, as the neural network, a multi-layer network represented by a convolutional neural network (CNN: Convolutional Neural Network), a multi-layer convolutional neural network (DCNN: Deep CNN), ResNet (Residual Network), DenseNet, or the like may be used.

The trained model 23 may be generated for each similar group based on the geological information of the construction end location. For example, the selection unit 115 selects data having similar geology as learning data. That is, the selection unit 115 prepares in advance the geological information and the excavation measurement result regarding the construction end location having similar geological tendencies (geological similarity, for example, the similarity of the column chart is equal to or higher than the threshold value). do. By machine learning the network model 20 using the learning data of each of the plurality of groups, the trained model 23 is generated for each tendency of the column chart. The learning data is not limited to the columnar chart, and may be grouped according to the similarity of the N values, or may be grouped according to the combination of the columnar chart and the N value, and other geology. It may be grouped based on information.

Further, the selection unit 115 may prepare the learning data by dividing the learning data into a plurality of groups of similar position information, that is, geological information regarding the construction end location near the address and the excavation measurement result. By training the model in the same manner using the training data of each of the plurality of groups grouped with similar position information, the trained model 23 for each area is generated.

Of course, the learning data may be prepared by grouping according to the combination of the geological information and the location information. When the learning data is grouped according to the combination of the geological information and the position information, the learning data may be selected by weighting the geological information and the position information, which parameter is prioritized.

The learning data may be created according to the planned construction site. As a method of selecting learning data, for example, two methods can be mentioned.

As the first selection method, the selection unit 115 inputs the geological information of the construction end place where the similarity of the columnar chart obtained from the geological information of the planned construction place is equal to or more than the threshold, although the place (position information) is different. 21 is set, and the measured integrated current value at the construction end location is selected as the correct answer data 22.

As a second method, the selection unit 115 selects the geological information of the construction end place within a certain range from the planned construction place as the input data 21 and the measured integrated current value of the construction end place as the correct answer data 22. do. When the second method is adopted, a threshold value may be set stepwise based on a certain range of distance from the planned construction site. Specifically, the selection unit 115 collects the geological information of the construction end place within a radius of 100 m from the planned construction place as input data, but the number of data of the geological information within the radius of 100 m is not sufficient (that is, the number of data). Is less than a predetermined number). In this case, the selection unit 115 is selected until the number of data becomes a predetermined number or more, such as geological information of the construction end place within a radius of 200 m, geological information of the construction end place within a radius of 500 m, and so on. The data collection range of the construction end location may be expanded.

Next, the time of using the model of the information processing apparatus according to the present embodiment will be described with reference to the conceptual diagram of FIG.
At the time of use, the model execution unit 113 of the information processing apparatus 1 inputs the processing target data including the position information and the geological information of the planned construction site into the trained model 23 as the input data 31, so that the expected excavation measurement result can be obtained. , The assumed integrated current value (hereinafter, also referred to as the assumed integrated current value) is generated as the output data 32. As the geological information of the planned construction site, for example, a columnar chart (geological data) in the depth direction and an N value are assumed, but the present invention is not limited to this.

Next, the operation of the information processing apparatus 1 according to the present embodiment will be described with reference to the flowchart of FIG.
In step S401, the acquisition unit 111 acquires the position information and the geological information of the planned construction site as the processing target data.

In step S402, the selection unit 115 selects the trained model corresponding to the data to be processed. For example, select a trained model of a group similar to the location and / or geological information of the data to be processed.

Specifically, if the trained model 23 is generated for each trend of the log, that is, for each similar geology, the similarity of the log will be based on the geological information (column) included in the processing target data. The trained model 23 trained by the training data of the group having the threshold value or more may be selected, and a suitable assumed integrated current value can be obtained. Further, if a trained model is generated for each area, the trained model has been trained based on the training data of the group within a certain range from the position indicated by the position information based on the position information included in the input data 31. A suitable assumed integrated current value can be obtained by selecting a model. Further, if the trained model is generated based on the combination of the position information and the geological information, the trained model based on the combination of the position information and the geological information of the processing target data may be selected.

In step S403, the model execution unit 113 inputs the data to be processed into the trained model selected in step S402, and the assumed integrated current value is generated.
In step S404, the display control unit 114 graphs the assumed integrated current value and the integrated current value obtained during excavation at the planned construction site and displays them on a display or the like.
If one trained model is generated, the process of step S402 may be omitted.

Further, as another example of application of the trained model, instead of preparing the trained model in advance, the trained model corresponding to the processed target data may be generated after the processing target data is acquired.

Another example of the operation of the information processing apparatus 1 according to the present embodiment will be described with reference to the flowchart of FIG.
In step S501, the acquisition unit 111 acquires the data to be processed in the same manner as in step S401.

In step S502, the selection unit 115 selects learning data based on the data to be processed. As the method for selecting the learning data, the method for selecting the learning data described above in FIG. 2 may be used. For example, learning data is selected in which the geological information of the construction end location within a certain range from the position information of the processing target data is used as input data and the actual excavation measurement result of the construction end location is used as the correct answer data.

In step S503, the learning unit 112 learns the network model using the learning data selected in step S502, and generates a trained model.
In step S504, the model execution unit 113 inputs the processing target data into the generated trained model and generates an assumed integrated current value.
In step S505, the display control unit 114 graphs the assumed integrated current value and the integrated current value obtained during excavation at the planned construction site and displays them on the display, as in step S404.

Next, a display example of the output data of the trained model will be described with reference to FIGS. 6A and 6B.
Here, an example displayed when the planned construction site is under construction is shown. As shown in FIG. 6A, the columnar diagram 61, the N value 62, and the integrated current value 63 are displayed in parallel on the display window 60 displayed on the display device. In the graph area of the integrated current value 63, a graph of the assumed integrated current value generated by the trained model and a graph of the integrated current value during excavation of the planned construction site are drawn in real time. Here, the assumed integrated current value is shown by a thin line, and the integrated current value being measured is shown by a thick line. Any display form may be used as long as the current values can be distinguished from each other.
An example is shown in which the assumed integrated current value and the integrated current value being measured are superimposed and displayed on one graph. As shown in FIG. 6B, the assumed integrated current value 64 and the integrated current value 65 being measured are parallel to each other. It may be displayed. When displaying in parallel, the heights of the same depth positions of both integrated current value graphs are arranged so as to match. In this way, the data are displayed in parallel so that they can be seen at a glance.

The graph data in which the columnar chart 61, the N value 62, and the assumed integrated current value are displayed in parallel may be displayed on the display device in advance or may be output to the outside by printout or the like.

According to the embodiment shown above, a trained model is generated in which the position information and the geological information regarding the construction end location are used as input data and the integrated current value when the construction end location is excavated is learned as the correct answer data. By inputting the position information and geological information of the planned location into the trained model, the assumed integrated current value can be generated. Then, the builder or the manager can predict the depth to reach the support layer based on the assumed integrated current value by creating the display window 60 as shown in FIGS. 6A and 6B.
Moreover, since the assumed integrated current value and the current value acquired in real time can be compared, it is possible to immediately determine the arrival of the support layer.
Then, since a large amount of the current value at the construction end point is used as the teacher data, the trained model can be generated accurately.
Furthermore, since the trained model is highly accurate, it is possible to accurately calculate the assumed current value for the planned construction site.
Further, since the accuracy of the assumed current value is high, the user can accurately evaluate the ground while observing the current value displayed by the display control unit.

Next, another embodiment will be described with reference to FIGS. 7 and 8.
As shown in FIG. 7, after performing steps S401 to S404 in the same manner as in the above-described embodiment, test excavation at the planned construction site is performed in step S405.
In step S406, the ground is excavated with the assumed integrated current value and the integrated current value during excavation displayed on the display in the test excavation, and the graph shape of the integrated current value during the excavation is the graph shape of the assumed integrated current value. The determination process of whether or not it is similar to is performed. The degree of similarity of the graph shape may be determined by the builder or the manager, or the degree of similarity may be determined by the processing circuit 11 of the information processing apparatus 1 or the like.
In particular, when it is assumed that the depth of the support layer set by the design or the like has been reached, it is confirmed whether or not the graph shape of the integrated current value during excavation behaves in the same manner as the graph shape of the assumed integrated current value.
Then, when it is determined that the graph shape of the integrated current value during excavation is similar to the graph shape of the assumed integrated current value, the trained model may be used as it is in the main construction of step S408.

On the other hand, if it is determined that the graph shape of the integrated current value during excavation does not resemble the graph shape of the assumed integrated current value, the geological information and integrated current value acquired by the test excavation are used as shown in FIG. The trained model is regenerated in step S407 as input data and correct answer data, respectively.
Even if it is determined that the graph shape of the integrated current value during excavation is similar to the graph shape of the assumed integrated current value, step S407 may be performed if it is desired to improve the accuracy of the trained model.
As shown in FIG. 7, in step S403 again, the processing target data is input to the trained model regenerated by the model execution unit 113, and the assumed integrated current value is regenerated.
In this embodiment, when it is determined that the graph shape of the assumed integrated current value and the graph shape of the integrated current value during test excavation are not similar, the integrated current value acquired at the time of test excavation has been learned as correct answer data. Since the model is modified, a highly accurate trained model can be generated.
Even if it is determined that the graph shape of the assumed integrated current value and the graph shape of the integrated current value during test excavation are similar, the trained model is regenerated using the integrated current value acquired during test excavation as correct data. Therefore, it is possible to generate a trained model with higher accuracy.

The instructions given in the processing procedures shown in each of the above embodiments can be executed based on a program that is software. It is also possible for a general-purpose computer system to store this program in a recording medium in advance and read the stored program to obtain an effect similar to the effect of the above-mentioned identification device. Further, the recording medium in the present embodiment is not limited to a medium independent of a computer or an embedded system, but also includes a recording medium in which a program transmitted by a LAN, the Internet, or the like is downloaded and stored or temporarily stored.

1 ... Information processing device, 11 ... Processing circuit, 12 ... Storage unit, 13 ... Communication interface, 20 ... Network model, 21,31 ... Input data, 22 ... Correct answer Data, 23 ... Trained model, 31 ... Input data, 32 ... Output data, 60 ... Display window, 61 ... Column diagram, 62 ... N value, 63 ... Integration Current value, 111 ... acquisition unit, 112 ... learning unit, 113 ... model execution unit, 114 ... display control unit, 115 ... selection unit.

Claims (10)

The acquisition department that acquires geological information about the planned construction site,
The construction schedule acquired by the acquisition unit for the trained model in which the geological information regarding the construction end location is used as input data and the current value indicating the excavation resistance when excavating the construction end location is learned as correct answer data. A model execution unit that generates an assumed current value by inputting the geological information regarding the location,
An information processing device including a display control unit for displaying the assumed current value generated by the model execution unit on a display device. The information processing apparatus according to claim 1, wherein the correct answer data is data relating to a construction end location having a geology having a geology having a similarity equal to or higher than a threshold value with the geology of the planned construction location. The information processing device according to claim 1 or 2, wherein the correct answer data is data relating to a construction end location whose distance from the planned construction location is within a threshold value. The information processing apparatus according to any one of claims 1 to 3, wherein the display control unit displays the assumed current value and the current value obtained during excavation at the planned construction site in parallel. .. The information processing apparatus according to any one of claims 1 to 3, wherein the display control unit superimposes and displays the current value and the current value obtained during excavation at the planned construction site. Computer,
Acquisition means to acquire geological information about the planned construction site,
The construction schedule acquired by the acquisition means for the trained model in which the geological information regarding the construction end location is used as input data and the current value indicating the excavation resistance when excavating the construction end location is learned as correct answer data. A model execution method that generates an assumed current value by inputting geological information about the location,
An information processing program for functioning as a display control means for displaying the assumed current value generated by the model execution means on a display device. A storage unit that stores construction data in which geological information regarding the construction end location and a current value indicating excavation resistance when excavating the construction end location are associated with each other.
A learning unit that trains a model using learning data that uses the geological information about the construction end location as input data and the current value when excavating the construction end location as correct data, and generates a trained model. A learning device comprising. The acquisition department that acquires the geological information about the planned construction site,
Further, the storage unit is further provided with a selection unit for selecting construction data related to a construction end location similar to the planned construction location as the learning data.
The learning device according to claim 7, wherein the learning unit trains a model using the learning data selected by the selection unit and generates the trained model. The learning device according to claim 8, wherein the selection unit selects construction data of a construction end location having a geology whose similarity with the geology of the planned construction location is equal to or higher than a threshold value. The learning device according to claim 8 or 9, wherein the selection unit selects construction data at a construction end location whose distance from the planned construction location is within a threshold value.

Images