Determination device

Abstract

A determination device (10) according to the present disclosure is provided with: a clustering unit (152) that selects one element as a target element from sensing data comprising a plurality of elements belonging to any of a plurality of orthogonal components, and clusters the selected target element and a plurality of elements in the vicinity of the target element; a calculation unit (153) that calculates a variance of the plurality of clustered elements for each component direction of the plurality of components; and a determination unit (154) that determines a component to which the target element belongs on the basis of the calculation result of the variance.

Description

Judgment device

[0001] The present disclosure relates to a determination device.

[0002] Non-Patent Document 1 describes a technique for determining the bearing capacity of deteriorated concrete structures by performing structural analysis. Deterioration of concrete structures includes spalling or cracking of concrete, corrosion of reinforcing bars, etc. Traditionally, analytical models for structural analysis have basically been created by manually inputting information about deterioration, but in recent years, with the improvement in the calculation capabilities of computers, research into automating structural analysis has been progressing.

[0003] G. Sibenika, I. Kovacica and V. Petrinas, Automated Model Preprocessing for Structural Analysis, International Symposium on Automation and Robotics in Construction (ISARC), 2021.

[0004] The rebars that make up concrete structures include rebars called main rebars and distribution rebars, which are arranged perpendicularly. Main rebars and distribution rebars have different functions. Main rebars are designed to resist tensile forces generated by loads applied to concrete members. Distribution rebars are designed to increase the shear strength of concrete members. Because of these differences in their functions, the corrosion of main rebars and distribution rebars affects the strength of concrete structures differently. Therefore, when automatically inputting the location of deteriorated rebars into an analytical model for structural analysis, it is necessary to automatically determine whether the rebars are main rebars or distribution rebars. In other words, a method is needed that can determine the component to which each of multiple elements (e.g., the location of deterioration of rebars) belongs, which belong to multiple orthogonal components (e.g., main rebars and distribution rebars).

[0005] In view of the above circumstances, an object of the present disclosure is to provide a determination device capable of determining the component to which each of a plurality of elements belonging to any of a plurality of orthogonal components belongs.

[0006] In order to solve the above problem, the determination device disclosed herein includes a clustering unit that selects one element as a target element from sensing data consisting of multiple elements that belong to any of multiple orthogonal components, and clusters the selected target element and multiple elements in the vicinity of the target element; a calculation unit that calculates the variance of the clustered multiple elements for each component direction of the multiple components; and a determination unit that determines the component to which the target element belongs based on the calculation result of the variance.

[0007] According to the determination device according to the present disclosure, it is possible to determine the component to which each of a plurality of elements belonging to one of a plurality of orthogonal components belongs.

[0008] Fig. 2 is a diagram showing an example of the configuration of a determination device according to the present disclosure. Fig. 3 is a diagram showing an example of the arrangement of main reinforcements and distribution reinforcements. Fig. 4 is a diagram for explaining clustering of a plurality of elements by the clustering unit shown in Fig. 1. Fig. 5 is a diagram for explaining calculation of variance for each component direction by the calculation unit shown in Fig. 1. Fig. 6 is a flowchart showing an example of the operation of the determination device shown in Fig. 1. Fig. 7 is a diagram for explaining setting of the number of elements to be clustered by the clustering shown in Fig. 1.

[0009] FIG. 1 is a diagram illustrating an example configuration of a determination device 10 according to an embodiment of the present disclosure. The determination device 10 according to the present disclosure determines the component to which each of a plurality of elements belonging to one of a plurality of orthogonal components belongs. In the following, an example will be described in which the plurality of orthogonal components are main reinforcement 1 and distribution reinforcement 2 in a concrete structure, as shown in FIG. 2. In the following, as shown in FIG. 2, it is assumed that the main reinforcement 1 extends in the x direction and the distribution reinforcement 2 extends in the y direction orthogonal to the x direction. Note that the plurality of components may also be underground pipes or the like buried in the orthogonal direction.

[0010] As shown in FIG. 1, a determination device 10 according to this embodiment includes an input unit 11, an output unit 12, a storage unit 13, a communication unit 14, and a control unit 15.

[0011] The input unit 11 includes one or more input interfaces that receive input operations from a user and acquire information based on the user operations. For example, the input unit 11 may be, but is not limited to, a pointing device, a keyboard, a mouse, or the like.

[0012] The output unit 12 includes one or more output interfaces that output information. For example, the output unit 12 is, but is not limited to, a display that outputs information as a video or a speaker that outputs information as an audio. The output unit 12 outputs the determination result of the determination device 10 as a video or an audio. If the output unit 12 is a touch panel display, it also functions as the input unit 11.

[0013] The storage unit 13 is one or more memories. The memory may be, for example, a semiconductor memory, a magnetic memory, or an optical memory, but is not limited to these, and may be any memory. The storage unit 13 is, for example, built into the determination device 10, but may also be configured to be accessible from outside the determination device 10 via any interface.

[0014] The communication unit 14 is configured to include one or more communication modules. The communication unit 14 may include, for example, a communication module compatible with a mobile communication standard such as 4G (4th Generation) or 5G (5th Generation). The communication unit 14 may include, for example, a communication module compatible with a wireless LAN standard (e.g., IEEE 802.11). The communication unit 14 may also include, for example, a communication module compatible with a wired LAN standard.

[0015] The control unit 15 is one or more processors. The processor may be, for example, a general-purpose processor or a dedicated processor specialized for a specific process, but is not limited to these and may be any processor. The control unit 15 controls the overall operation of the determination device 10.

[0016] The determination device 10 has the following software configuration: One or more programs used to control the operation of the determination device 10 are stored in the storage unit 13. When the program stored in the storage unit 13 is read by the control unit 15, it causes the control unit 15 to function as an acquisition unit 151, a clustering unit 152, a calculation unit 153, and a determination unit 154.

[0017] The acquisition unit 151 acquires sensing data input via the input unit 11 or transmitted from an external device via the communication unit 14. The sensing data is discrete data consisting of multiple elements belonging to any of multiple known orthogonal components. As shown in FIG. 2 , when deterioration (corrosion) occurs in the reinforcing bars (main reinforcement 1 and distribution reinforcement 2) of a concrete structure, rust fluid is exposed on the concrete surface near the reinforcing bars due to the corrosion of the reinforcing bars. Therefore, for example, by analyzing a photographed image of the surface of the concrete structure and detecting the exposed location of rust fluid, the location of the deteriorated reinforcing bars can be identified. The sensing data is discrete data consisting of multiple elements, such as pixel values ​​at the deteriorated location of the reinforcing bars identified in a photographed image of the surface of a concrete structure in which the main reinforcement 1 and distribution reinforcement 2 (multiple elements) are arranged.

[0018] The clustering unit 152 selects one element as a target element from the sensing data acquired by the acquisition unit 151, which is made up of multiple elements belonging to any of multiple orthogonal components, as shown in Fig. 3. The clustering unit 152 clusters the selected target element and multiple elements in the vicinity of the target element. In Fig. 3, the elements within the dotted circle are shown as having been clustered.

[0019] The clustering unit 152 clusters multiple elements using, for example, a clustering method using machine learning. One clustering method is the k-nearest neighbor method, which selects k elements neighboring a target element. In the k-nearest neighbor method, the number k of neighboring points (elements) to be clustered must be set manually. In contrast, in this embodiment, the clustering unit 152 automatically sets k. Methods for automatically setting the number k of elements to be clustered include the elbow method and the silhouette analysis method. However, these methods are intended to evaluate the accuracy of the clustering results, and are not intended to set a value of k suitable for determining the component to which an element belongs, as in the present disclosure. Details of the method for setting k by the clustering unit 152 will be described later.

[0020] The calculation unit 153 calculates the variance of each component direction of multiple components of multiple elements (target element and multiple elements in the vicinity of the target element) clustered by the clustering unit 152. That is, as shown in Fig. 4, the calculation unit 153 calculates the variance σx of the component direction (x direction) of the main reinforcement 1 and the variance σy of the component direction of the distribution reinforcement 2 of the clustered elements.

[0021] The determination unit 154 determines the component to which the target element belongs based on the calculation result of the variance for each component direction by the calculation unit 153. The determination unit 154 determines the component to which the target element belongs, for example, by comparing the magnitude of the variance for each component direction or by principal component analysis. When determining the component to which the target element belongs by comparing the magnitude of the variance for each component direction, the determination unit 154 determines that the target element belongs to the component (i.e., the reinforcing bar 2) corresponding to the component direction with the larger variance (the y direction in the example shown in FIG. 4 ).

[0022] Next, the operation of the determination device 10 according to this embodiment will be described with reference to a flowchart shown in FIG.

[0023] The acquisition unit 151 acquires sensing data made up of a plurality of elements belonging to any of a plurality of orthogonal components (step S101).

[0024] The clustering unit 152 selects one element as a target element from the sensing data (step S102). The clustering unit 152 clusters the target element and multiple elements in its vicinity using a clustering method using machine learning in accordance with a set hyperparameter (step S103). The hyperparameter is, for example, the number k of elements to be clustered.

[0025] The calculation unit 153 calculates the variance of the clustered elements in each of the component directions (in this embodiment, the variance σx in the x direction and the variance σy in the y direction) (step S104).

[0026] The clustering unit 152 calculates the ratio cur_ratio (=σy / σx) of the variances calculated for each component direction (step S105).

[0027] The clustering unit 152 determines whether or not the maximum value of the ratio of variances for each component direction (maximum ratio max_ratio) has been set (step S106).

[0028] If it is determined that the maximum ratio max_ratio has not been set (step S106: No), the clustering unit 152 proceeds to the process of step S108, which will be described later.

[0029] If it is determined that the maximum ratio max_ratio has already been set (step S106: Yes), the clustering unit 152 determines whether cur_ratio>max_ratio (step S107).

[0030] If it is determined that cur_ratio>max_ratio is satisfied (step S107: Yes), the clustering unit 152 sets the ratio cur_ratio of variances for each component direction to the maximum ratio max_ratio (step S108).Then, the clustering unit 152 changes the hyperparameters (step S109) and returns to the processing of step S103.

[0031] If it is determined that cur_ratio>max_ratio is not true (that is, cur_ratio≦max_ratio) (step S107: No), the clustering unit 152 sets the value of the hyperparameter to the previous value (step S110).

[0032] By the processes of steps S103 to S110, while cur_ratio>max_ratio, the ratio of variance for each component direction calculated each time the hyperparameter is updated is set to the maximum ratio max_ratio. Then, when cur_ratio≦max_ratio, the hyperparameter is set to the previous value.

[0033] That is, as shown in FIG. 6 , the clustering unit 152 changes the number k of elements to be clustered, and sets the value of k when the variance ratio for each component direction becomes a maximum value as the number k of elements to be clustered. Note that FIG. 6 illustrates an example in which the variance ratio for each component direction has a maximum value, but this is not limiting. Depending on the definition of the variance ratio for each component direction, the variance ratio for each component direction may have a minimum value. In this case, the clustering unit 152 may set the value of k when the variance ratio for each component direction becomes a minimum value as the number k of elements to be clustered. In this way, the clustering unit 152 changes the number k of elements to be clustered, and sets the number k of elements to be clustered so that the variance ratio cur_ratio for each component direction becomes a maximum or minimum value. It is considered that the variance ratio cur_ratio for each component direction becomes a maximum or minimum value when the difference in variance for each component direction is the largest. In such a case, by determining the component to which the target element belongs, it is possible to improve the accuracy of the determination.

[0034] Referring again to FIG. 5, the clustering unit 152 clusters the target element and a plurality of elements in the vicinity of the target element using a set number k (step S111).

[0035] The calculation unit 153 calculates the variance of the clustered elements for each component direction of the multiple components (step S112).

[0036] The determining unit 154 determines the component to which the target element belongs based on the variance calculation result.

[0037] As described above, the determination device 10 according to this embodiment includes a clustering unit 152, a calculation unit 153, and a determination unit 154. The clustering unit 152 selects one element as a target element from sensing data consisting of multiple elements belonging to any of multiple orthogonal components, and clusters the selected target element and multiple elements in the vicinity of the target element. The calculation unit 153 calculates the variance of the clustered multiple elements for each component direction of the multiple components. The determination unit 154 determines the component to which the target element belongs based on the variance calculation result.

[0038] If a target element belongs to one of multiple components, it is highly likely that elements near the target element also belong to that component. If the number of elements belonging to one component is greater than the number of elements belonging to other components, the variance in the component direction of that component will be greater than the component directions of the other components. Therefore, the component to which the target element belongs can be determined based on the calculation results of the variance of the clustered elements for each of the multiple components.

[0039] By determining the component to which the target element belongs, it is possible to understand the state of each component. In addition, by determining the component to which the target element belongs, preprocessing for detailed structural analysis can be performed automatically, leading to reduced inspection and repair costs.

[0040] As described above, the program that controls the overall operation of the determination device 10 may be stored in a computer-readable storage medium. Using such a storage medium, the program can be installed on a computer. Here, the storage medium on which the program is stored may be a non-transitory storage medium. The non-transitory storage medium is not particularly limited, and may be, for example, a CD-ROM, a DVD-ROM, or a USB (Universal Serial Bus) memory. Furthermore, the program may be downloaded from an external device via a network.

[0041] The following additional notes are provided regarding the above-described embodiments.

[0042] [Supplementary Item 1] A determination device comprising a control unit, wherein the control unit is configured to: select one element as a target element from sensing data consisting of a plurality of elements belonging to any of a plurality of orthogonal components; cluster the selected target element and a plurality of elements in the vicinity of the target element; calculate the variance of the clustered elements for each component direction of the plurality of components; and determine the component to which the target element belongs based on the calculation result of the variance.

[0043] [Supplementary Item 2] In the determination device according to Supplementary Item 1, the control unit changes the number of the elements to be clustered and sets the number of the elements to be clustered so that a ratio of variance for each of the component directions becomes a maximum value or a minimum value.

[0044] [Supplementary Item 3] A determination method executed by a determination device, comprising the steps of: selecting one element as a target element from sensing data consisting of a plurality of elements belonging to any of a plurality of orthogonal components; and clustering the selected target element and a plurality of elements in the vicinity of the target element; calculating the variance of the clustered elements for each component direction of the plurality of components; and determining the component to which the target element belongs based on the calculation result of the variance.

[0045] [Supplementary Item 4] A non-transitory storage medium storing a program executable by a computer, the non-transitory storage medium storing the program causing the computer to operate as the determination device according to Supplementary Item 1.

[0046] All publications, patent applications, and technologies mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, and technology was specifically and individually indicated to be incorporated by reference.

[0047] Although the above-described embodiments have been described as typical examples, it will be apparent to those skilled in the art that many modifications and substitutions can be made within the spirit and scope of the present disclosure. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications or alterations are possible without departing from the scope of the claims.

[0048] REFERENCE SIGNS LIST 10 Determination device 11 Input unit 12 Output unit 13 Storage unit 14 Communication unit 15 Control unit 151 Acquisition unit 152 Clustering unit 153 Calculation unit 154 Determination unit

Claims

1. A determination device comprising: a clustering unit that selects one element as a target element from sensing data consisting of multiple elements that belong to one of multiple orthogonal components, and clusters the selected target element and multiple elements in the vicinity of the target element; a calculation unit that calculates the variance of the clustered multiple elements for each component direction of the multiple components; and a determination unit that determines the component to which the target element belongs based on the calculation result of the variance.

2. A determination device according to claim 1, wherein the clustering unit changes the number of elements to be clustered and sets the number of elements to be clustered so that the ratio of variance for each component direction becomes a maximum or minimum value.

Images