Stress monitoring device, stress monitoring system and monitoring system

Abstract

The system has an information acquisition unit that acquires measurement information of deformation or torsion applied to a structural member forming a structure, and a memory unit that stores linked stress information formed by linking the measurement information with design information of the structure and placement information of the structural member linked to the design information, and enables the linked stress information to be output to a terminal.

Description

Technology Field

[0001] The present invention relates to a stress monitoring device, etc., for acquiring information on the stress distribution applied to a structure. Background Technology

[0002] Conventionally, methods for measuring stress distribution in structures such as bridges or highways are known, and one such method is the infrared thermography method using an infrared camera. In the infrared thermography method, minute temperature fluctuations, or thermoelastic temperature fluctuations, that occur in an object during elastic deformation are measured by an infrared camera (infrared sensor), and defects in the object are detected based on this. In addition, another method is known in which the space between bridge piers is photographed by an infrared camera, the amount of temperature change is calculated based on the thermal image data captured when a vehicle passes, and the amount of stress change is calculated based on the amount of temperature change to measure the stress distribution occurring in the bridge (see, for example, Patent Document 1). Prior art literature

[0003] International Publication No. 2017 / 141294 The problem to be solved

[0004] In infrared thermography, while it is possible to measure the stress distribution on the surface of a structure within the range captured by an infrared camera, it is necessary to deploy multiple infrared cameras to determine the stress distribution across the entire structure. This is particularly true for large structures, where multiple infrared cameras must be positioned around the structure. Furthermore, when a structure is located outdoors, it is difficult to determine the exact stress distribution due to the temperature difference between sunny and shaded areas. Additionally, climatic conditions, such as exposure to wind and rain, can cause the temperature to drop, making it difficult to accurately determine the stress distribution. Consequently, determining stress distribution based on temperature changes cannot be performed accurately depending on the installation location of the measurement target; this leads to problems such as the overlooking of serious abnormalities like cracks, or the inability to determine the degree of fatigue or endurance limits.

[0005] The present invention aims to provide a means for acquiring high-certainty stress information of the whole or part of a structure. means of solving the problem

[0006] The stress monitoring device of the present invention comprises an information acquisition unit for acquiring measurement information of deformation or torsion applied to a structural member forming a structure, and a memory unit for storing linked stress information linked to the structural member's placement information, which includes the measurement information, design information of the structure, and placement position information of the structural member that can be linked to the design information, and is capable of outputting the linked stress information to a terminal. In addition, the design information of the structure and / or the placement information of the structural member may be acquired in advance. Furthermore, the placement information of the structural member may be configured to include placement posture information of the structural member with respect to the structure. In addition, the structural member may be a joining member that is joined to each other across two or more structural members.

[0007] In addition, the stress monitoring device of the present invention has a calculation unit that calculates stress distribution information acting on the whole or part of the structure from the linked stress information, and enables the stress distribution information calculated by the calculation unit to be substituted for the linked stress information or output together with the linked stress information to the terminal.

[0008] In addition, the stress monitoring device of the present invention generates a machine learning model for calculating stress distribution information from linked stress information by the calculation unit, and has a learning unit that manages the parameters of the machine learning model, an analysis unit that interprets the linked stress information to acquire analysis stress distribution information applied to the stress distribution of the structure, and a judgment unit that compares the stress distribution information by the machine learning model with the analysis stress distribution information by the analysis unit and determines the degree of agreement. The learning unit generates a machine learning model that updates the parameters by performing machine learning on the analysis stress distribution information as teaching data to improve the degree of agreement by the judgment unit.

[0009] In addition, the stress monitoring device of the present invention enables the output of the stress distribution information together with, or instead of, the linked stress information when the degree of agreement determined by the judgment unit is greater than or equal to a threshold.

[0010] In addition, the stress monitoring device of the present invention has a learning unit that generates a deep learning model.

[0011] In addition, the stress monitoring device of the present invention comprises an analysis unit that inversely analyzes the linked stress information to determine analysis conditions based on the result of the inverse analysis, and performs a forward analysis using the finite element method with the analysis conditions and the design information of the structure to obtain the analysis stress distribution information based on the result of the forward analysis.

[0012] In addition, the stress monitoring device of the present invention includes load conditions in the analysis conditions.

[0013] In addition, the stress monitoring device of the present invention acquires the measurement information by means of a torsion sensor installed on the structural member by the information acquisition unit.

[0014] In addition, the stress monitoring device of the present invention acquires the measurement information by means of a torsion sensor installed on a joining member disposed at a location where the information acquiring unit joins the members constituting the structure.

[0015] In addition, the stress monitoring device of the present invention is disposed across two joined members to which the torsion sensor is fixed.

[0016] In addition, the stress monitoring device of the present invention has the torsion sensor forming the joint member type.

[0017] In addition, the stress monitoring device of the present invention has the torsion sensor integrally formed with the structural member.

[0018] In addition, the stress monitoring device of the present invention has the joining member formed in any one of the following shapes: male screw type, female screw type, washer type, socket type, tubular type, plate type, L-shape, or T-shape.

[0019] In addition, the stress monitoring device of the present invention is configured such that the structure is formed by joining two or more members.

[0020] In addition, the stress monitoring system of the present invention comprises the stress monitoring device and a torsion sensor disposed on the joining member and connected to the stress monitoring device, and is configured to output various information from the stress monitoring device to an external terminal.

[0021] In addition, the monitoring system of the present invention comprises an information acquisition unit that acquires measurement information according to a target; an inverse analysis unit that performs inverse analysis on the measurement information to estimate a cause factor and its level; a forward analysis unit that performs forward analysis using the cause factor and its level estimated by the inverse analysis unit and acquires a state of the target according to the factor and its level based on the result of the forward analysis; a calculation unit that calculates a virtual state of the target using the measurement information; a learning unit that generates a machine learning model for calculating the virtual state by the calculation unit and manages the parameters of the machine learning model; and a judgment unit that compares the state acquired by the forward analysis unit with the virtual state and determines the degree of agreement. The learning unit performs machine learning on the state of the target as teacher data to update the parameters so that the degree of agreement by the judgment unit is improved. Effects of the invention

[0022] According to the present invention, high-certainty stress information can be obtained from the measurement information of a torsion sensor. Brief explanation of the drawing

[0023] [Fig. 1] This is a block diagram showing a stress monitoring system of the present embodiment. [Fig. 2] This is a block diagram showing a stress monitoring device. [Fig. 3] This is a drawing showing a bolt-type sensor. [Fig. 4] This is a block diagram showing the configuration of a circuit board. [Fig. 5] This is a flowchart showing the information transmission processing in a stress monitoring system according to the present embodiment. [Fig. 6] This is a drawing showing a structure. [Fig. 7] This is a diagram showing a part of the data constituting the coupled stress information. [Fig. 8] This is a drawing showing the load conditions required by inverse analysis. [Fig. 9] This is a diagram showing the measurement result image. [Fig. 10] This is a diagram showing a stress distribution image. Specific details for implementing the invention

[0024] An embodiment of the stress monitoring system of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a stress monitoring system of the present embodiment. The stress monitoring system (1) is configured to connect a stress monitoring device (2) and a user terminal (4) through a network (6). A plurality of bolt-type sensors (40) (sensors of a joining member type) are connected to the stress monitoring device (2) so as to be able to communicate through a dedicated line, etc.

[0025] The bolt-type sensor (40) forms a bolt shape according to structural members used in a structure composed of two or more members joined together, such as a structure (including construction, buildings, etc.), a vehicle (means of transportation), a home appliance, an agricultural machine, a construction machine (including general construction machine or special construction machine), or a machine tool. Additionally, the bolt-type sensor (40) can detect stress generated by a load. That is, the bolt-type sensor (40) functions as a sensor in the form of a bolt (joining member) that is joined together across two or more structural members of the structure.

[0026] The loads referred to here include single loads, repetitive loads, alternating loads, and impact loads; of course, in addition to traffic loads, collision loads, seismic loads, wind loads, thermal loads, snow loads, and water loads, loads caused by hydrostatic pressure, water flow, and buoyancy are also included.

[0027] Here, FIG. 2 is a block diagram showing a stress monitoring device (2). The stress monitoring device (2) is equipped with a CPU (20), and the CPU (20) is connected to a communication unit (22), an analysis unit (24), a memory unit (26), a machine learning unit (28), a learning memory unit (30), a distribution information creation unit (32), and a design data storage unit (36).

[0028] The communication unit (22) performs communication processing such as transmitting linked stress information (described later) or virtual stress distribution information (described later) to the user terminal (4) and receiving measurement information from the bolt-type sensor (40). The analysis unit (24) performs an analysis (inverse analysis) to specify load conditions based on stress information acting on a plurality of bolt-type sensors (40) for each structure, or a forward analysis (FEM analysis) using the finite element method to create analysis stress distribution information for the entire structure based on load conditions, etc.

[0029] The analysis stress distribution information obtained by the analysis of the analysis unit (24) includes information such as the stress distribution applied to the entire structure, and shape changes caused by twisting or bending of the structure.

[0030] The memory unit (26) stores instructional data consisting of information representing the stress distribution state applied to a part or the whole of the structure when a load is applied to the structure. Here, the stress distribution is determined by load conditions such as the position, direction, orientation, and magnitude of the load applied to the structure. Therefore, in the present invention, the analysis stress distribution information obtained by the analysis unit (24) is treated as instructional data, and the memory unit (26) stores analysis stress distribution information corresponding to various load conditions for each structure as instructional data.

[0031] The machine learning unit (28) performs machine learning and generates a neural network model (machine learning model) for creating a stress distribution of a structure from stress information of a bolt-type sensor (40). In addition, the machine learning unit (28) performs operations such as learning weight parameters from teacher data accumulated in the memory unit (26) in the learning unit of the neural network model. Furthermore, the machine learning model is not limited to a neural network model, but may also be a logistic regression model, a deep learning model, etc.

[0032] The learning memory unit (30) stores information according to machine learning models, such as different neural network models and weight parameters, for each structure in which the bolt-type sensor (40) is used. The distribution information creation unit (32) creates virtual stress distribution information representing the stress distribution of the entire structure from the result of inputting the stress information of the bolt-type sensor (40) into the neural network model.

[0033] In addition, the stress information includes, in addition to the stress value itself, dimensional information, shape information, material information, etc. of the bolt-type sensor (40), and includes deformation or torsion amounts that can calculate the stress value. Furthermore, the stress information may also include electrical resistance values, voltage values, or current values ​​that can be converted into deformation or torsion amounts or stress values.

[0034] The design data storage unit (36) is a database that manages design data by associating the component information of the component constituting the structure, the structure information (e.g., CAD data, design information), and the bolt information of the bolt-type sensor (40) used in the structure with the structure ID, which is the identifier of the structure.

[0035] Here, the member information may include information such as the physical properties of each part specified by the material of each part constituting the structure (for example, in the case where the structure is a building, the roof, exterior wall, floor, wall, ceiling, interior window, etc.), the shape of each part, and the method of joining (connecting) the parts to each other.

[0036] Additionally, the bolt information is information composed of bolt placement information, bolt material information, shape information, and dimensional information, etc., linked to the sensor ID assigned to each bolt-type sensor (40). The placement information is information indicating the placement location or placement posture of the bolt-type sensor (40). That is, the placement information includes information indicating where each bolt-type sensor (40) is placed in the structure and in what posture (for example, a posture in which the axial direction is horizontal (or vertical) and facing south).

[0037] The user terminal (4) is an information processing terminal that displays an image of stress acting on a structure having a bolt-type sensor (40). The user terminal (4) may be able to receive various information from the stress monitoring device (2) through the network (6), and the types thereof are not limited.

[0038] That is, the user terminal (4) is, for example, a smartphone, a mobile phone (feature phone), a PDA (Personal Digital Assistant), a wearable terminal (head-mounted display, glasses-type device, etc.), a tablet terminal, a notebook PC, a desktop PC, and various other computers or monitors having computing circuits.

[0039] The network (6) can be either a wireless network or a wired network. Specifically, it can be a wireless LAN, a Wide Area Network (WAN), ISDNs (Integrated Service Digital Networks), LTE (Long Term Evolution), LTE-Advanced, CDMA (Code Division Multiple Access), a 5th generation mobile communication system (5G), LPWA (LAW Power Wide Area), etc. Of course, the network can be Wi-Fi, a public switched telephone network, Bluetooth, an optical line, an ADSL (Asymmetric Digital Subscriber Line) line, a satellite communication network, etc., or a combination of these.

[0040] FIG. 3 is a drawing showing a bolt-type sensor (40). The bolt-type sensor (40) shown in this embodiment has a head (42) and a shaft (44) (of course, the head is not mandatory). The head (42) is installed at one end of the shaft (44) and has an outer diameter shape that is larger than the shaft (44). The shaft (44) has a threaded portion (44a) having a male screw groove formed on the tip side and a circumferential portion (44b) of a non-threaded area located between the head (42) and the threaded portion (44a) (of course, the circumferential portion (44b) is not mandatory).

[0041] The head (42) has a circuit board (50) installed on its front surface, and a torsion measurement sensor part (60) (described later) formed on the surface of the shaft part (44) is connected to the circuit board (50). Additionally, the head (42) may be equipped with an unillustrated cover that covers at least the front circuit board (50).

[0042] The screw portion (44a) of the shaft portion (44) has a first male screw spiral structure that forms a spiral groove in a predetermined lead angle and / or lead direction, and a second male screw spiral structure in which a spiral groove is set in a different lead angle and / or lead direction from the lead angle and / or lead direction of the first male screw spiral structure.

[0043] Here, two types of male screw spiral structures are formed overlappingly on the same area in the axial direction of the bolt-type sensor (2): a first male screw spiral structure that is configured to be a right screw, in which a female screw-type spiral groove formed as a corresponding right screw is configured to be interlocked, and a second male screw spiral structure formed as a left screw, in which a female screw-type spiral groove formed as a corresponding left screw is configured to be interlocked. Of course, the first male screw spiral structure and the second male screw spiral structure may be set so that the lead direction of the right screw is the same and the lead angles are different. Additionally, although the spiral grooves do not necessarily have to be formed in an overlapping manner, it is desirable to have a mechanism that can suppress loosening as a joining member for performing torsional measurement and stress measurement with high precision and reliability.

[0044] In this embodiment, the screw portion (44a) can be engaged with either a female screw body of a right screw or a left screw. Additionally, regarding the details of the screw portion (44a) having two types of male screw spiral grooves formed therein, Japanese Registered Patent Publication No. 4663813 according to the inventor of the present invention is referenced.

[0045] The circumferential portion (44b) has a conduction path formed directly on its outer surface that serves as a pattern for measuring torsion, and the torsion measuring sensor portion (60) is installed directly on the circumferential portion (44b). Additionally, the circumferential portion (44b) has an outer diameter less than or equal to the curve diameter or effective diameter of the screw portion (44a).

[0046] As a method of directly forming a torsion measurement sensor part (60) on the outer surface of the cylindrical part (44b), for example, an electrical insulation layer is directly formed on the outer surface by lamination printing, putty printing, coating, plating, inkjet printing, etc., and a torsion measurement sensor (60) is directly formed thereon. Additionally, the method of forming the electrical insulation layer is not limited to each of the above methods, and various methods may be employed, such as forming a film by sputtering an insulating material while a predetermined mask is placed, applying a silica material and heat treating it, or applying an organic insulating material such as a silicon-based, polyimide-based, epoxy-based, or urethane-based material.

[0047] Additionally, the torsional measurement sensor (60) is formed directly on an electrical insulating layer by lamination printing using a conductive paste, putty printing, coating, plating, inkjet printing, sputtering, etc. Electrical contacts are formed at both ends of the torsional measurement sensor (60) and connected to a circuit board (50), thereby allowing voltage to be applied between the electrical contacts.

[0048] As a result, the twisting occurring in the circumferential portion (44b) and the deformation of the circumferential portion (44b) can be detected by the change in resistance value of the torsion measurement sensor portion (60) to which voltage is applied to the electric contact plate. Additionally, if the shaft portion (44) (circular portion (44b)) is made of a non-conductive material, the torsion measurement sensor portion (60) may be formed directly on the outer surface of the circumferential portion (44b). Furthermore, a coating layer with excellent wear resistance, damage resistance, heat resistance, moisture barrier properties, solvent resistance, gas barrier properties, and deformation resistance (adhesion) may be formed to cover the torsion measurement sensor portion (60).

[0049] FIG. 4 is a block diagram showing the configuration of a circuit board (50). A control circuit (52) that controls each part collectively is arranged on the circuit board (50), and a measurement memory unit (53), a measurement unit (54), an RF (Radio Frequency) unit (56), a power circuit (58), etc. are connected to the control circuit (52). The measurement memory unit (53) can store a sensor ID or the elongation elasticity modulus of the shaft part (44), and can also store measurement information based on the measurement results by the measurement unit (54).

[0050] The measuring unit (54) is equipped with a measuring means for measuring stress information based on changes in electrical resistance detected through the torsional measuring sensor unit (60). Additionally, the measuring unit (54) may further include a temperature measuring means, a magnetic measuring means, an acceleration measuring means, and a gyro sensing means capable of measuring angular acceleration based on Coriolis's law. Accordingly, the measuring unit (54) can output measuring information including stress information and measurement results from other measuring means.

[0051] The RF unit (56) transmits the measurement information output by the measurement unit (54) to the stress monitoring device (2) through an antenna not shown.

[0052] The power circuit (58) uses a battery, etc., as a power source, converts the voltage to a voltage for driving each part such as the control circuit (52), and supplies power. Additionally, the battery may be, for example, a lithium battery or a small all-in-one battery that can be mounted on the circuit board (50), but a secondary battery may also be used. Additionally, if a secondary battery is used, a power receiving means (not shown) may be connected to the power circuit (58), and the power converted by the power receiving means may be supplied to the battery.

[0053] The receiving means converts vibrations, temperature differences, external light (lighting light or natural light), pressure changes (atmospheric pressure changes), external force changes (wind, buoyancy changes, rain), electromagnetic waves, etc., into electrical power. That is, the receiving means may have, for example, a coil, an antenna, a light receiving part (photonic element field component, solar panel, etc.), an energy conversion element such as a vibrator configured to convert vibrations or pressure changes, etc., into electrical energy such as a microphone or a piezoelectric element, or a Seebeck element or circuit utilizing the Seebeck effect. Additionally, it goes without saying that the battery may be wired to the power circuit (58) through an external power supply (AC adapter) as a commercial power source.

[0054] Additionally, the control circuit (52) is preferably used to control each part to operate continuously, but in the case where a secondary battery is used, in order to suppress power consumption, the timing of operation of each part, such as measurement by the measurement unit (54) and transmission by the RF unit (56), is adjusted.

[0055] The operation timing can be appropriately set, such as by setting based on time or by setting based on the amount of charge. Also, it is not necessary to operate each part simultaneously, and the RF part (56) may be controlled to operate at a lower frequency compared to the measurement part (54). For example, the measurement part (54) may be controlled to operate at a predetermined measurement interval (e.g., every hour), and the RF part (56) may be controlled to operate at a predetermined transmission interval (e.g., every day).

[0056] Additionally, the RF unit (56) may be controlled to operate whenever the measurement unit (54) performs a predetermined number of measurements (e.g., 5 times). In this way, when the interval between the operation of the measurement unit (54) and the operation of the RF unit (56) is different, the control circuit (52) associates the measurement time obtained by an unillustrated timing function with the measurement information from the measurement unit (54) and stores it in the measurement memory unit (53). Then, when the RF unit (56) operates, the measurement information is read from the measurement memory unit (53) and transmitted to the stress monitoring device (2).

[0057] In addition, the acceleration measuring means or gyro sensing means of the measuring unit (54) may be operated at all times, and the operation timing may be set so that the stress information is measured when the acceleration measuring means or gyro sensing means detects a large change greater than a predetermined value.

[0058] Next, with reference to the flowchart shown in FIG. 5, the information transmission processing by the stress monitoring system (1) according to the present embodiment will be described. Here, the structure is described as an example of a rigid frame structure having two columns (100a, 100b) and a beam (110) as shown in FIG. 6. Also, the control circuit (52) is configured to keep each part constantly operating.

[0059] The columns (100a, 100b) of the structure are fixed by being embedded in the ground, and a column support metal fitting (not shown) for vertically fixing the columns (100a, 100b) is fixed by two bolt-type sensors (40a, 40b) in the shape of anchor bolts. In addition, a main beam (110) is placed between the columns (100a, 100b) and is fixed to the columns (100a, 100b) by two bolt-type sensors (40c, 40d).

[0060] That is, the bolt-type sensor (40a) is installed vertically in the axial direction and is used to fix the column (100a) to the ground, the bolt-type sensor (40b) is installed vertically in the axial direction and is used to fix the column (100b) to the ground, the bolt-type sensor (40c) is installed horizontally in the axial direction and is used to fix the column (100a) and the beam (110), and the bolt-type sensor (40d) is installed horizontally in the axial direction and is used to fix the column (100b) and the beam (110).

[0061] Additionally, the stress monitoring device (2) performs constant monitoring, and when a predetermined load is applied to the structure, each control circuit (52) of the bolt-type sensors (40a~40d) acquires measurement information by the measurement unit (54) and transmits the sensor ID and measurement information to the stress monitoring device (2) by the RF unit (56).

[0062] The CPU (10) of the stress monitoring device (2) receives sensor IDs and measurement information from each bolt-type sensor (40a~40d) through the communication unit (22) (Step S1). The CPU (10) reads design data by referring to the design data storage unit (36) based on the sensor ID. That is, it reads design data containing bolt information with the same sensor ID as the received sensor ID from the design data storage unit (36). Then, the CPU (10) creates linked stress information by linking the received measurement information (stress information) with the design data.

[0063] FIG. 7 is a diagram showing a part of the data constituting the linked stress information, and the linked stress information is composed of the sensor ID of each bolt-type sensor (40a~40d), stress as measurement information, and design data (discharge location information, discharge posture information, structure ID, etc.) being linked together. Additionally, although not shown, other information included in the design data (structure information, member information, etc.) is also part of the information constituting the linked stress information. The linked stress information is stored in the memory unit (26).

[0064] As shown in FIG. 7, each stress based on measurement information was a stress value of 577 N / mm² that could be linked to the bolt-type sensor (40a) of sensor ID (001), a stress value of 194 N / mm² that could be linked to the bolt-type sensor (40b) of sensor ID (002), a stress value of 638 N / mm² that could be linked to the bolt-type sensor (40c) of sensor ID (003), and a stress value of 671 N / mm² that could be linked to the bolt-type sensor (40d) of sensor ID (004).

[0065] The CPU (10) performs inverse analysis processing by the analysis unit (24) using the design data and the stress values ​​of each bolt-type sensor (40a~40d) among the linked stress information (step S2). In this inverse analysis processing, load conditions acting on the structure (magnitude of the load, location and direction where the load is applied, etc.) are specified. That is, when the above stress values ​​are applied to each bolt-type sensor (40a~40d) in the structure determined by the design data, the load conditions of the load that can be applied to the structure are specified by inverse analysis. The method of inverse analysis processing is not particularly limited and is appropriately set from well-known inverse analysis methods.

[0066] The CPU (10) stores specific load conditions in the memory unit (26) by reverse analysis by the analysis unit (24). The load conditions specified here are such that the position of the load shown in FIG. 8 is at the top position of the column (100a), the direction of the load is approximately horizontal from the side of the column (100a) toward the side of the beam (110), and the magnitude of the load is 100 kN.

[0067] Next, the CPU (10) sets the analysis conditions for the FEM analysis as a forward analysis (step S3). Specifically, the load conditions and design data (member information, structure information, bolt information) are used as the analysis conditions. The CPU (10) executes the FEM analysis by the analysis unit (24) to generate analysis stress distribution information (step S4). That is, the analysis unit (24) inputs a load according to the load conditions into an FEM analysis model that is approximately equivalent to an actual structure that can be reproduced by the design data, and outputs analysis stress distribution information as the result of the analysis of the stress distribution of the entire structure.

[0068] The CPU (10) can store the analysis stress distribution information, structure ID, sensor ID, measurement information, load conditions, etc., as teaching data in the memory unit (26). Additionally, the teaching data is not limited to the above configuration, and can be appropriately set, for example, a configuration consisting of analysis stress distribution information, structure ID, and measurement information, or a configuration consisting of analysis stress distribution information, sensor ID, measurement information, and load conditions.

[0069] Additionally, the CPU (10) inputs the stress values ​​of the bolt-type sensors (40a to 40d) and design data into a neural network model, and creates virtual stress distribution information applied to the entire structure from the output result (Step S5). Specifically, the CPU (10) uses the design data read after Step S1. The CPU (10) inputs the measurement information and design data into the neural network model stored in the learning memory unit (30) and obtains an output result.

[0070] The output result obtained at this time is a calculation result according to the neural network model, and the distribution information creation unit (32) creates virtual stress distribution information from the output result. In addition, the CPU (10) may perform the process of creating virtual stress distribution information in step S5 in parallel with the process of creating analysis stress distribution information represented by steps S2 to S4.

[0071] Next, the CPU (10) compares the analysis stress distribution information with the virtual stress distribution information (step S6) and determines whether the degree of agreement between the stress distributions is greater than or equal to a threshold (e.g., 70%) (step S7). The determination of the degree of agreement between the stress distributions may be performed by comparing the stress distribution of the entire structure, but to reduce the processing burden, it may also be performed by specifying multiple parts (members) in the structure and comparing the stress distributions in each specified part. Additionally, parts subject to comparison may include locations in the structure that receive load and are important for maintaining the shape of the structure, locations prone to change (failure or deformation, etc.) due to stress, and locations prone to stress application.

[0072] When the degree of agreement of the stress distribution is below the threshold (step S7, No), the CPU (10) performs machine learning (step S8). That is, after providing the teacher data stored in the memory unit (26), the machine learning unit (28) learns the weight parameters of the neural network model so that the stress distribution of the virtual stress distribution information matches the stress distribution of the analysis stress distribution information.

[0073] The CPU (10) updates the weight parameters stored in the learning memory unit (30) according to the processing result by the machine learning unit (28).

[0074] Additionally, the CPU (10) transmits linked stress information to the user terminal (4) via the network (6) by means of the communication unit (22) (step S9) and terminates the information transmission process. The user terminal (4) that receives the linked stress information displays a measurement result image consisting of an overall view of the structure, the position and orientation of the bolt-type sensors (40a~40d) shown within the overall view, and stress values ​​as measurement results from each bolt-type sensor (40a~40d), as shown in FIG. 9. In addition, stress values ​​are displayed in the measurement result image, but are not limited thereto and may display measurement information such as torsion amount or deformation amount.

[0075] When the degree of agreement of the stress distribution is greater than the threshold (step S7, Yes), the CPU (10) transmits virtual stress distribution information as high-likelihood stress distribution information to the user terminal (4) via the network (6) by the communication unit (22) (step S10) and terminates the information transmission process. Additionally, the CPU (10) performs an update of weight parameters based on machine learning even when the degree of agreement of the stress distribution is greater than the threshold. As a result, the degree of agreement of the stress distribution is further improved, and it becomes possible to create virtual stress distribution information (high-likelihood stress distribution information) with higher accuracy among high-likelihood stress distribution information.

[0076] A user terminal (4) that receives high stress distribution information displays a stress distribution image created according to the high stress distribution information. In the stress distribution image, as shown in FIG. 10, a structure elastically deformed by a load and a stress distribution superimposed on the structure are shown. The stress distribution is represented by an intermittent or continuous gradient using colors, color shades, etc., and can indicate that the stress is smaller as the color of the gradient is lighter, and the stress is larger as the color is darker.

[0077] Accordingly, as shown in FIG. 10, the gradient color can be displayed more intensely in the area around the structure where the columns (100a, 100b) are bent to the right, the bending point in the columns (100a, 100b) of the structure, or the bolt-type sensor (40c, 40d) (see FIG. 8) with a high stress value of measurement information, while the gradient color can be displayed less intensely in the area near the ground of the columns (100a, 100b) that is far from the location where the load condition is applied and far from the bending point, making it possible to display the appearance of the stress distribution in a way that is easy to understand visually and intuitively.

[0078] Additionally, the display content of the stress distribution image is not limited to the above, and the display content can be appropriately set, such as the position of the bolt-type sensors (40a to 40d) within the structure, the stress value based on the measurement information of each bolt-type sensor (40a to 40d), the position of the load based on the load condition, the direction of the load, the magnitude of the load, etc.

[0079] Therefore, the user can intuitively verify the appearance of the structure under load and the stress distribution through the stress distribution image. Additionally, the user can verify the appearance of the structure even from a distance from the monitored structure.

[0080] As described above, according to the stress monitoring system (1) of the present invention, stress distribution information is generated based on load conditions obtained by inverse analysis, and by transmitting the stress distribution information, a stress distribution image is displayed on a user terminal (4). Therefore, the user can check the stress distribution or deformation of the structure (structure) from the stress distribution image even from a location far from the structure. Furthermore, the user can identify the condition of the structure or where the stress is concentrated from the stress distribution image, and can determine the soundness, such as the presence or absence of yielding, the degree of fatigue, and the endurance limit. In addition, even in a location where the structure cannot be seen, it is possible to identify the location where abnormalities such as fine cracks may occur.

[0081] In addition, since the virtual stress distribution information transmitted to the user terminal (4) represents a stress distribution approximately identical to the analysis stress distribution information obtained by FEM analysis as a pure analysis, it can represent a high-likelihood stress distribution with high accuracy for the user, and more preferably, the highest-likelihood stress distribution information with the highest likelihood among the high-likelihood stress distribution information.

[0082] In addition, even if the stress distribution of the virtual stress distribution information differs from the stress distribution of the analytical stress distribution information, machine learning is performed to update the weight parameters of the neural network model, and the stress distribution of the virtual stress distribution information is gradually approached to match the stress distribution of the analytical stress distribution information more closely, thereby improving the accuracy of the stress distribution based on the virtual stress distribution information.

[0083] When transmitting virtual stress distribution information to a user terminal (4), it is desirable to output a result that is, for example, completely identical to the analysis stress distribution information and is calculated as the most possible maximum likelihood stress distribution. Therefore, it is desirable for machine learning to continue running until virtual stress distribution information in which the maximum likelihood stress distribution appears is created.

[0084] In addition, since linked stress information can be transmitted even if the degree of agreement between virtual stress distribution information and analytical stress distribution information is below a threshold, the stress applied to the bolt-type sensors (joining members) is displayed on the user terminal; consequently, the user can identify the stress state acting on each bolt-type sensor within the structure and determine structural integrity, such as estimating locations where stress is concentrated.

[0085] In addition, the stress monitoring device automatically performs inverse analysis, FEM analysis as forward analysis, processing of virtual stress distribution information, and calculation of load conditions including load information acting on the structure, automatic creation of supervised data, updating of weight parameters of the neural network model, and output of high-probability or maximum-probability stress distributions through artificial intelligence algorithms based on deep learning (machine learning) via deep learning. In other words, since the stress monitoring device, which is equipped with an artificial intelligence program, performs various processes through autonomous control, manual control by an administrator or other personnel becomes unnecessary. Furthermore, if machine learning can be optimized based on the results of machine learning on various structures, it becomes possible to significantly reduce the number of supervised data required to output the results of the high-probability stress distribution for each structure, as well as the time required for machine learning.

[0086] In addition, in the above-described embodiment, the creation of analysis stress distribution information is performed whenever measurement information is acquired due to the accumulation of teacher data. However, when virtual stress distribution information (maximum stress distribution information) corresponding to the above-described maximum stress distribution is created, the creation of analysis stress distribution information, etc., for that structure may not be performed, and the virtual stress distribution information may be transmitted to the user terminal. That is, when the creation of maximum stress distribution information is approximately certain, it is considered that the accumulation of teacher data is sufficient, and the processing by the analysis unit (24) and the comparison processing between the analysis stress distribution information and the virtual stress distribution information may not be performed, and the virtual stress distribution information may be transmitted to the user terminal. By doing so, the processing burden of the stress monitoring device (2) can be reduced.

[0087] In addition, when it is generally certain that the virtual stress distribution information generated by the neural network model is actual stress distribution information, the comparison processing between the analysis stress distribution information and the virtual stress distribution information for the structure may be omitted, and the virtual stress distribution information may be transmitted to the user terminal as actual stress distribution information. This can also reduce the processing burden of the stress monitoring device (2). However, due to the accumulation of training data, the processing of generating analysis stress distribution information, i.e., training data, and the execution of machine learning processing by the machine learning unit continue.

[0088] In this way, if some processing such as processing by the analysis unit (24) or comparison processing is omitted, the processing time required from the acquisition of measurement information to the transmission of virtual stress distribution information is significantly reduced, and as a result, the latest stress distribution image can be displayed on the user terminal immediately, so that safety can be verified immediately when a load is applied to the structure, thereby improving convenience.

[0089] In addition, in the above-described embodiment, although it was explained that the RF unit (56) transmits the sensor ID and measurement information to the stress monitoring device (2) through an antenna, it is not limited thereto, and the control circuit (52) and the stress monitoring device (2) may be wired to enable communication so as to transmit the sensor ID and measurement information. Alternatively, a relay device may be connected to the stress monitoring device (2), and the relay device may relay the sensor ID and stress value from the RF unit (56) and transmit them to the stress monitoring device (2).

[0090] In addition, although the case of using a male screw-type bolt sensor as a joining member for joining two or more members of a structure was described as an example, the shape of the joining member is not limited to a male screw type, and any shape that is used for joining two or more members, such as a female screw type, washer type, socket type, tube type, plate type, L-shape, T-shape, etc., may be used.

[0091] In addition, although the case of measuring stress from the torsion of a bolt-type sensor installed in the structure was described as an example, it is, of course, not limited to this; a torsion measuring sensor may be formed on the structural member itself constituting the structure, and the stress based on the torsion of the member may be measured directly.

[0092] In addition, it is preferable that the structural member on which the torsion measurement sensor is installed be positioned at a location where two or more structural members are joined together. Referring to FIGS. 9 and 10, it can be seen that stress is concentrated in the bolt-type sensor (40c, 40d) positioned between the column (100a, 100b) and the main beam (110) as joined members. This is because, among the structural members, stress can be significantly applied to the joining member that fixes the structural members that are joined together. Therefore, in monitoring the stress of the structure, it can be said that it is desirable to measure the stress of the structural member (joining member) for joining the structural members together.

[0093] In addition, although the above-described embodiment was explained as an example where only the case where a load is applied, it is also possible to install a temperature measuring means in the measuring unit (54) and perform temperature correction of the amount of twisting or to specify the thermal load from the temperature of the bolt-type sensor measured by the temperature measuring means. Therefore, analysis conditions can be set considering the thermal load caused by the rise or fall of ambient temperature and the temperature difference between the inside and outside of the structure.

[0094] In addition, the measuring unit (54) may measure vibration information of the structure caused by an earthquake using an acceleration measuring means or a gyro sensing means. In this way, it is possible to determine from which direction the shaking of the structure was transmitted, and also to reliably detect the tilt even if the tilt is slight, in the case where the entire structure is tilted due to ground subsidence caused by an earthquake, etc.

[0095] In addition, in the above-described embodiment, the case in which virtual stress distribution information or linked stress information generated by the stress monitoring device is merely transmitted to the user terminal was described as an example, but it is not limited thereto. For example, when the stress monitoring device receives a request from the user terminal, it may output virtual stress distribution information or linked stress information.

[0096] In addition, if virtual stress distribution information or linked stress information output from a stress monitoring device can be received at the user terminal side, the stress monitoring device may create and store the virtual stress distribution information or linked stress information, and the user terminal may access the stress monitoring device via a network to download (streaming may also be used) the latest virtual stress distribution information or linked stress information of a predetermined structure.

[0097] In addition, the coupled stress information or virtual stress distribution information (high-risk stress distribution information, maximum-risk stress distribution information) may be information intended to display a stress distribution image or a measurement result image on a user terminal, and the data format is not specifically limited. That is, the coupled stress information and virtual stress distribution information may be generated as image data.

[0098] Additionally, the information, such as linked stress information and virtual stress distribution information output by the stress monitoring device, may include text data, table data, graph information, as well as still image information, moving image information, etc. Furthermore, in order to facilitate understanding of changes in stress distribution over time, it is desirable to use visual information, particularly moving image information. The creation of moving images can be appropriately set, for example, by creating multiple stress distribution images as still images (for example, images equivalent to one minute as a video) and then displaying the multiple stress distribution images on a user terminal.

[0099] Furthermore, if the processing time from the acquisition of measurement information to the transmission of virtual stress distribution information can be significantly reduced, it becomes easier to display changes in the stress distribution acting on the structure in real time as a moving image. Specifically, the stress distribution image may be displayed to a user terminal via streaming transmission, and in this way, the stress distribution that changes over time can be displayed to the user terminal.

[0100] In addition, the structure utilizing the bolt-type sensor in the above-described embodiment is not particularly limited and may include various objects (vehicles, structures, etc.), devices (home appliances, etc.), and machinery (agricultural machinery, construction machinery, machine tools, etc.). For example, vehicles or mobile bodies include manned or unmanned movement, and for land use, include multi-wheeled vehicles and wheelless vehicles. Multi-wheeled vehicles with four or more wheels include trucks, buses, fire trucks, aerial ladder trucks, pump trucks, trailers, tank trucks, mixer trucks, crane trucks, street sweepers, vacuum trucks, tow trucks, snowplows, subways, monorails, Shinkansen, trams, locomotives, cable cars, armored vehicles, tanks, etc. Four-wheeled vehicles include automobiles, police cars, ambulances, campervans, etc. Three-wheeled vehicles include tricycles and automatic tricycles. Two-wheeled vehicles include motorcycles, scooters, bicycles, electric assist bicycles, wheelchairs, strollers, rickshaws, etc. Single-wheeled vehicles include single-wheeled vehicles. Wheelless vehicles include linear motor cars, hovercrafts, elevators, escalators, lifts, gondolas, ropeways, Ferris wheels, carousels, swings, seesaws, etc. Other wheeled vehicles include wheel loaders, tire rollers, load rollers, graders, asphalt finishers, motor sweepers, dump trucks, wheel cranes, forklifts, straddle carriers, turret-type interior carriers, agricultural tractors, agricultural pesticide sprayers, harvesting and threshing vehicles, rice transplanters, combines, excavators, backhoes, loading shovels, snowmobiles, roller skates, roller shoes, skateboards, kickboards, skis, snowboards, skate shoes, etc.Airborne vehicles or mobile bodies include light aircraft such as balloons and airships, airplanes, rotary-wing aircraft, gliders, helicopters, radio-controlled unmanned aerial vehicles, rockets, and heavy aircraft such as vertical take-off and landing aircraft, regardless of whether they are manned or unmanned. Marine vehicles include boats, yachts, passenger ships, cargo ships, oil tankers, fishing vessels, warships, submarines, etc. In addition, parts used in the vehicle include a monocoque, body shell, bonnet, door, tailgate, front fender, radiator grille, bumper, meter, heater, windshield glass, door windglass, engine, radiator, muffler, brake pedal, accelerator pedal, clutch pedal, seat, exhaust pipe, tail lamp, head lamp, wheel, tire, track, propeller, tank, cylinder, piston, actuator, damper, linear guide, bearing, chassis, shaft, etc.

[0101] Structures include buildings and civil structures (works). Buildings include residential, commercial facilities, public facilities, cultural facilities, educational facilities, medical facilities, entertainment facilities, transportation facilities, industrial facilities, religious facilities, military facilities, and plant facilities. Buildings also include detached houses, attached houses (terrace houses), collective housing, multi-unit housing, share houses, office buildings, churches, monasteries, temples, shrines, castles, palaces, imperial palaces, gardens, parks, hospitals, clinics, train stations, historical sites, airports, government housing, government offices, police stations, fire stations, police substations, stadiums, sports fields, swimming pools, schools, gymnasiums, theaters, cinemas, drill grounds, entertainment venues, viewing areas, public halls, assembly halls, hotels, inns, factories, warehouses, logistics warehouses, logistics centers, boarding houses, dormitories, child welfare facilities, midwifery clinics, rehabilitation support facilities for the physically disabled, social reintegration facilities for the mentally disabled, protection facilities, women's protection facilities, support facilities for the intellectually disabled, and elderly welfare facilities. Facilities include paid nursing homes, special care nursing homes, day service centers, maternal and child health facilities, movie theaters, museums, art galleries, libraries, public restrooms, sports practice ranges, bowling alleys, ski resorts, department stores, supermarkets, general stores, exhibition halls, cabarets, cafes, nightclubs, bars, dance halls, playgrounds, public baths, waiting rooms, restaurants, eateries, shops, car garages, car repair shops, movie studios, television studios, etc.

[0102] Civil engineering structures (works) include bridges, metal structures, railways, roads, ports, coasts, rivers, power generation facilities or equipment, dams, tunnels, land improvement structures, disaster prevention structures, and agricultural civil engineering structures.Bridges include girder bridges, cable-stayed bridges, truss bridges, arch bridges, Rhamen bridges (or rigid frame bridges), and suspension bridges; metal structures include tower structures, storage structures, sluice gates / locks, hydraulic iron pipes, composite structures, water supply systems, sewage systems, and various pipelines including goods for the delivery of gas or oil; railways include tracks, track structures, roadbeds, railway stations, signal / security / communication facilities, high-speed railways, special railways, cableways, and urban railways; roads include roadbeds / roadbeds, pavements, asphalt pavements, concrete pavements, ballast roads, dust-proof roads, etc.; and ports include anchorages, breakwaters, revetments, Includes revetments, jetties, docks, quays, piers, temporary buildings, cargo handling and onshore facilities, ship and car communication facilities, fishing ports, navigational aids, etc.; coasts include coastal structures; rivers include revetments and revetments, sand arrestation, river structures, and canals; power generation facilities or equipment include water intake facilities, reservoirs, regulating reservoirs, nuclear power plants, thermal power plants, hydroelectric power plants, tidal power, geothermal power, wave power, and wind power; dams include gravity dams, fill-type dams, and arch dams; tunnels include tunnel structures; and land improvement structures include land It includes formation, land reclamation, dredging, reduction, drainage, land clearing, and mixing of soils.

[0103] In addition, members used in structures include stone materials including natural stone, artificial stone, and crushed stone; timber; pre-cut timber; natural wood; machine-grade wood; visual-grade wood; ungraded timber; Class A structural material; Class B structural material; laminated timber; engineered wood; structural laminated timber; plywood; wood products; resin products; metal products; steel; steel plates; steel frames; organic materials; coating materials; waterproofing materials; cement paste; slag; mortar; concrete; tile; porcelain tile; stoneware tile; earthenware tile; tatami; lightweight cellular concrete; Konpane; water-resistant plywood; gypsum board; and fireproofing. Includes boards, calcium silicate boards, insulation materials, glass fibers, rock wool, rigid urethane foam, styrofoam, phenolic foam, polystyrene foam, cellulose fibers, prestressed concrete, precast concrete, underwater concrete, polymer concrete, resin concrete, mass concrete, expansive concrete, low-shrinkage concrete, and non-shrinkage concrete.

[0104] Parts of a structure include structural materials, fastening members, finishing materials, foundation materials, interior materials, exterior materials, facing materials, insulation materials, foundation, base, wall, column (column, tubular column, corner post, splash post, major pillar), main beam (roof beam, dragon beam, round beam, ascending beam), roof, ceiling, floor, stairs, puncheon, vertical roof puncheon, window, glass window, window frame, door, door frame, front door, shelf, Geonmok, doorcase (upper frame of doorcase), decorative pillar, alcove, decorated rail of alcove, floor board of alcove, alcove post, transom, lintel, attached lintel, girth, dragon beam, braced frame, gwan, stud, puncheon, eaves, eaves soffit, ceiling joist Includes awnings, gutters, flooring, carpets, cushioned floors, cloth, wallpaper, sliding doors, window paper, siding, expansion joints, roof tiles, cement tiles, slate tiles, colonial tiles, corrugated steel plates, and handrails.

[0105] The structure in which each member of the building is realized includes masonry, brickwork, timber, wood structure, soil structure, steel structure (S structure), light steel structure (LGS), heavy steel structure, plain concrete structure, reinforced concrete structure (RC structure), steel reinforced concrete structure (SRC structure), concrete-filled steel tube structure, cement block structure (CB structure), reinforced cement block structure, steel-concrete composite structure, prestressed concrete structure (PC), membrane structure, wall structure, frame structure, masonry structure, air membrane structure (single membrane), air membrane structure (double membrane), air membrane structure (air beam), pure rigid frame structure, walled rigid frame, and strake rigid frame. Includes frame), brace structure, core structure, tube structure, truss structure, bolt structure, shell structure, cable (hanging) structure, pin structure, space frame, arch, dome shell, earthquake resistant structure, base isolation structure, vibration damping structure, rigid structure, flexible structure, and earthquake-proof structure.

[0106] Home appliances include video devices (display devices) such as televisions and projectors; video devices (recording and playback devices) such as video tape recorders, DVD recorders, Blu-ray Disc recorders, HDD recorders, DVD players, and Blu-ray Disc players; video devices (shooting devices) such as video cameras and digital cameras; audio devices (recording and playback devices) such as wire recorders, tape recorders, mini-disc recorders, radio cassettes, and IC recorders; audio devices (playback devices) such as analog players, CD players, amplifiers, and radios; audio devices (reproduction devices) such as speakers and headphones; white goods (refrigerators, washing machines, etc.); and information appliances.

[0107] Agricultural machinery includes general-purpose agricultural machinery such as tractors; agricultural machinery used for culrivation and soil preparation such as plows, harrows, rollers, rotary tillers, puddling machines, land rollers, levelers, ridgers, and slitting machines; agricultural machinery used for soil cultivation, formation, and improvement such as rooting machines, pan breakers, grooving machines, mold rinas (culvert drilling machines), diggers, and backhoes; and manure spreaders (compost Machines used for fertilization such as compost fertilizer distributors, slurry spreaders, lime saws, planters (seed-shooting seeders), and fertilizer sowers; agricultural machinery used for sowing and transplanting such as rice transplanters, vegetable transplanters, transplanters (transplanting machines), and broadcasters; sprayers, power sprayers, misters, dusters, power dusters, granule applicators, power applicators, foggers, aerial sprayers / helicopters (aerial pest control), soil disinfectors, brushcutters, farm masters, and speed. Sprayer, frost damage protection machine, medium cultivating and weeding machine, thinner (intermediate machine,Agricultural machinery used for pest control and management, such as thinning machines, power pumps, and sprinklers (irrigation devices); agricultural machinery used for harvesting, such as binders, combines, vegetable harvesters, mowers, hay balers, roll balers, windrowers, threshing machines, bean cutters (legume harvesters), corn harvesters, corn shellers, potato harvesters, beet harvesters, sweet potato digging machines, sweet potato vine cutters, sugarcane harvesters, peanut harvesters, flax harvesters, onion diggers, chestnut threshers, peanut threshers, pluckers, mulberry shoot reaping machines, specialty crop diggers, vibratory harvesters, and hop flower thinning machines; dryers, rice hullers, and sorters. Agricultural machinery used for drying and preparing harvested products, such as rice milling machines, forage dryers, chicken manure dryers, specialty crop dryers, grain milling machines, disc mowers, mower conditioners, tedders, rakes, forage harvesters, hay balers, hay presses, load wagons, hay loaders, bale loaders, feed cutters, forage blowers, silage unloaders, feed grinders, feed choppers, root cutters, feed mixers, and feed molders; automatic feeders; milking machines; milk coolers; waterers; and water heaters. Manure spreader, livestock barn cleaner, solid-liquid separator, manure treatment device, heater,Agricultural machinery used for livestock management, such as egg lifters, sorting machines for cleaning eggs, and barn disinfectors; humidity and temperature control devices for the communal rearing of young silkworms, power mulberry feeders, automatic rearing machines for young silkworms, automatic rearing machines for grown silkworms, delimbing machines, cocoon harvesting machines, and floss removing machines; agricultural machinery used for sericulture, such as mulchers, crucific cultivating machines, greenhouse heaters, vegetable washers, deep fertilizer applicators, power pruning shears, tree towers, rotary cutters for orchards, and fruit sorting machines. This includes agricultural machinery used for vegetable and fruit horticulture (field farming) within agricultural land, such as transport containers and cultivators; agricultural machinery used for tea production, such as tea steaming machines, primary drying tea rollers, rolling freshly-picked tea-leaves machines, secondary drying tea rollers, and final drying tea rollers; agricultural machinery used for ornamental and specialty crops, such as pruning machines, ramie peelers, rush classifiers, tulip sorters, and peanut shellers; agricultural machinery used for forestry, such as bush cutters, chain saws, and yaders; and agricultural machinery used for transport and conveyance, such as trailers, grain conveyors, and front loaders.

[0108] General construction machinery includes bulldozers and scrapers such as ripper bulldozers (bulldozers with rip saws), scraper bulldozers, towed scrapers, and motor scrapers; hydraulic shovels (Yunbo, backhoe, power shovel); draglines, clamshells, on-mud excavators, tractor shovels; excavators and mucking shovels such as wheel loaders, trenchers, and bucket wheel excavators; trucks, dump trucks, trucks equipped with cranes; trailers; locomotives; muck cars; shuttle cars; and rough terrain haulers (specialized equipment Transport machinery such as special equipment haulers, shovel loaders, forklifts, all-wheel drive vehicles, belt conveyors, and bucket wheel excavators; crawler cranes, truck cranes, wheel cranes (all-terrain cranes, rough terrain cranes), tower cranes, jib cranes, railway cranes, floating cranes, pipe layers, construction lifts, elevators, portal cranes, forklifts, straddle carriers, container carriers, top lifters, clamp lifts, aerial vehicles (lift cars), concrete floor finishing robots, and sling-excluding robots.Cranes and cargo handling machinery such as unloaders, pile drivers, diesel hammers, hydraulic hammers, vibratory hammers, water jets (water jet cutters), earth augers, inner excavators of earth augers, drop hammers, hydraulic steel pipe press-in-and-drawing machines, sand pile drivers, powder injection agitators, all-casing excavators, hole digging and pole-standing cars, earth drills, reverse circulation drills, underground continuous wall construction machines, mud wastewater treatment facilities (including alkaline water neutralizers and sludge suction and discharge vehicles (vacuum cars)), grout Pump (grout pump), grout mixer (including mortar plant), pneumatic caisson construction equipment, deep mixing processing machine, ground improvement machine for high pressure injection stirring, chemical liquid injection and construction equipment, deep foundation work machineryFoundation construction machinery such as caisson pile machines (rotary injectors, underwater cutters, etc.), pile extractors, bowling machines, down-the-hole hammers, rock borers (hand hammers, leg hammers, drifters, pick hammers, baby hammers, riveting hammers, chipping hammers, caulking hammers, scaling hammers, sand rammers, concrete breakers, large breakers, etc.), drill jumbos, crawler drills, tunnel boring machines / cutters, grab hoppers, grab lifters, tunnel construction equipment, shield construction equipment, drilling machinery / tunnel construction machinery, motor graders, stabilizers, mixing plants, and mixing machines for ultra-soft ground Motor graders and subgrade machinery (sub-basic machines) such as (for super soft ground), road rollers (macadam rollers, tandem rollers), tire rollers, tamping rollers, vibratory rollers, tampers, rammers, compacting equipment such as vibrating compactors, concrete plants, truck mixers (agitator body trucks), concrete pump trucks, concrete pumps, concrete pressurers, screw cleats, agitator cars, concrete crushers, concrete machinery, asphalt plants, recycling plants, asphalt finishers, asphalt kettles, distributors,Chip spreader, asphalt cooker, concrete spreader, concrete finisher, concrete leveler (concrete vertical finishing machine), concrete simple finishing machine, concrete lateral yarding machine, vibrating joint cutting machine, concrete cutter, inner vibrator, asphalt engine sprayer, asphalt curbing machine, joint sealer, play susceptor, slip form paper, packaging machinery such as food preservation processing machines, road surface heater, joint cleaner, road sweeper, line marker, dissolving tank, lane marker removal machine, road surface cutting machine, road surface layer regenerating machine, guardrail sweeper, road surface safety groove cutting machine (grooving machine), Road maintenance machines such as sprinkler trucks, guardrail pole drive machines, compartment line construction machines, slab upper surface thickening machines, micro surface machines, and drainage pavement function recovery machines; air compressors (pressure), blowers (fans), etc., air compressors and blowers,Pumps such as small spiral pumps, small multistage centrifugal pumps (turbine pumps), underwater motor pumps for deep wells, vacuum pumps, construction underwater motor pumps (submersible pumps), underwater sand pumps (construction underwater pumps equipped with stirring devices), and slurry pumps; electrical equipment such as transformers, switches in high-pressure air, cubicle-type high-voltage substation facilities, and motor generators; winches, hoists, chain blocks, etc.; track scales, meters, core collectors (core boring machines), CBR testers, flat loading test apparatuses, grout flow and pressure measuring devices, gas detectors, sound meters, vibration measuring instruments, settling and inclination measuring equipment, and dust meters. Test and measuring instruments such as turbidimeters, automatic surveying instruments, and optical measuring instruments, erected girders, bentgrass, gantry cranes, hoists, chain blocks, geared trolleys, winches, jacks, hydraulic pumps, bogies for heavy weight, delivery devices, steel towers, carriers, saddles, backstay adjusting devices, cable anchoring devices, turnbuckles, rope hangers, unreelers, and delivery devices.Temporary facilities apparatus for installation work on steel bridges and PC bridges, such as lateral yarding machines, lowering devices, traveler cranes, girder hanging devices, girder hanging gate preparation moving devices, turntables, moving timber, and working cars of ground cover and handrails; concrete mixers, aggregate measuring instruments, concrete buckets, concrete vibrators, concrete crushers, jaw crushers, impact crushers, welding machines, welding rod dryers, hydraulic jacks, drop hammers, rails, turntables, mortar concrete spray machines, concrete sprayers, accelerating agent supply devices, seed spray machines, and bentonite. Bentonite mixer, water tank, bush cutter, lawn mower, chain saw, float, construction signal, construction high-pressure washer, chemical sprayer, hay collector, jet heater, garbage truck, self-propelled crusher, self-propelled soil improver,Includes other general land-based construction machinery such as self-propelled wood crushers.

[0109] In addition, the present invention can also be applied to special construction machinery, and special construction machinery includes, among other things, machinery for port, river, and coastal construction such as main work vessels, auxiliary work vessels, auxiliary equipment for work vessels, auxiliary equipment for port construction, concrete production equipment, concrete transport equipment, concrete cooling equipment, aggregate production equipment, cement transport and storage equipment, water supply equipment, pollution control equipment, dam construction machinery such as other dam construction machinery, snow removal machinery such as snowmobiles, snow removal devices, snow removal attachments, and spreading vehicles, sewage construction machinery such as propulsion construction machinery, mine removal machinery, etc.

[0110] Machine tools include lathes such as general-purpose or NC turret lathes, milling cutters, milling machines using end mills, shaping machines using bites, planning machines using bites, drilling machines using drills, reamers, taps, boring machines using bites, electrical discharge machining (EDM) machines such as wire-cut EDM machines and engraving EDM machines, broaching machines using broaches, gear hobbing machines (hobs), gear cutting machines such as gear shapers (rack cutters, pinion cutters), grinders using grinding wheels, contour machines, band sawing machines, and machining centers. It includes centers, water jet processing machines, laser processing machines, electron beam processing machines, honing machines, electrochemical processing machines, deburring and chamfering machines, electrochemical deburring machining machines, cutting machines, rolling machines, forming devices such as nut formers or part formers, etc. In addition, it includes various processing tools such as drills, end mills, bits, chips, taps, toothed tools, dies, molds, and tool holders included in these machine tools.

[0111] In addition, structural elements include structural materials for frames, structural materials for buildings, exterior materials, interior materials, propellers for wind power generation, solar panels, cylinders, pistons, actuators, dampers, linear guides, bearings, engine blocks, chassis, shafts, rods, L-shaped fittings, curtains, roller curtains, railings / lattices / sashes for verandas / balconies / outdoor windows, railings for corridors, passageways, stairs, bedsides, or bathroom interior walls, chairs, desks, bookshelves, cupboards, beds, bathtubs, toilet seats, toilets, etc.

[0112] In addition, although the above embodiment describes an example of monitoring stress, the system may not necessarily monitor stress. For example, it may monitor temperature distribution, heat conduction, vibration propagation status, etc.

[0113] In addition, the subjects of monitoring or analysis may include, in addition to the aforementioned structures, the human body, facial expressions, emotions, health conditions, weather, topography (faults, etc.), celestial bodies, images, sounds, etc. Explanation of the symbols

[0114] 1… measurement system, 2… stress monitoring device, 4… user terminal, 6… network, 20… CPU, 22… communication unit, 24… analysis unit, 26… memory unit, 28… machine learning unit, 30… learning memory unit, 32… distribution information creation unit, 36… design data storage unit, 40… bolt-type sensor, 42… head unit, 44… shaft unit, 44a… screw unit, 44b… circumference unit, 50… circuit board, 52… control circuit, 53… measurement memory unit, 54… measurement unit, 56… RF unit, 58… power circuit, 60… torsion measurement sensor.

Claims

Claim 1 A stress monitoring device characterized by comprising: an information acquisition unit for acquiring measurement information of deformation or torsion applied to a structural member forming a structure; a memory unit for storing linked stress information formed by linking the measurement information with design information and / or excretion information of the structure; a calculation unit for calculating stress distribution information acting on the whole or part of the structure from the linked stress information; a learning unit for generating a machine learning model to calculate stress distribution information from the linked stress information by the calculation unit and managing the parameters of the machine learning model; an analysis unit for acquiring analysis stress distribution information of the stress distribution of the structure by interpreting the linked stress information; and a judgment unit for comparing the stress distribution information by the machine learning model with the analysis stress distribution information by the analysis unit and determining the degree of agreement, wherein the learning unit performs machine learning on the analysis stress distribution information as teaching data to update the parameters so as to improve the degree of agreement by the judgment unit. Claim 2 delete Claim 3 delete Claim 4 A stress monitoring device according to claim 1, characterized in that when the degree of agreement by the judgment unit is greater than or equal to a threshold, the stress distribution information can be output. Claim 5 A stress monitoring device according to claim 1, wherein the learning unit generates a deep learning model. Claim 6 A stress monitoring device according to claim 1, wherein the analysis unit performs inverse analysis using the linked stress information to specify analysis conditions, performs forward analysis using the finite element method using the analysis conditions and design data of the structure, and acquires the analysis stress distribution information based on the result of the forward analysis. Claim 7 A stress monitoring device according to claim 6, wherein the above interpretation conditions include load conditions. Claim 8 A stress monitoring device according to claim 1, wherein the information acquisition unit acquires the measurement information by means of a torsion sensor installed on the structural member. Claim 9 A stress monitoring device according to claim 8, characterized in that the torsion sensor is integrally formed with the structural member. Claim 10 A stress monitoring device according to claim 8, characterized in that the torsion sensor is disposed on the structural member disposed across between two joined members fixed to each other. Claim 11 A stress monitoring device according to claim 8, wherein the torsion sensor is characterized by forming a structural member type. Claim 12 A stress monitoring device according to claim 8, wherein the structural member on which the torsion sensor is installed forms the shape of any one of the following joining members: male screw type, female screw type, washer type, socket type, tubular type, plate type, L-shaped type, or T-shaped type. Claim 13 A stress monitoring device according to claim 1, characterized in that the structure is formed by joining two or more structural members. Claim 14 A stress monitoring system characterized by having a stress monitoring device according to claim 1, a torsion sensor disposed on the structural member and connected to the stress monitoring device, and configured to output various information from the stress monitoring device to an external terminal. Claim 15 A monitoring system comprising: an information acquisition unit for acquiring measurement information according to a target; an inverse analysis unit for estimating a cause factor and its level by performing inverse analysis on the measurement information; a forward analysis unit for performing forward analysis using the cause factor and its level estimated by the inverse analysis unit and acquiring a state of the target according to the factor and its level based on the result of the forward analysis; a calculation unit for calculating a virtual state of the target using the measurement information; a learning unit for generating a machine learning model to calculate the virtual state by the calculation unit and managing the parameters of the machine learning model; and a judgment unit for comparing the state acquired by the forward analysis unit with the virtual state and determining the degree of agreement, wherein the learning unit performs machine learning on the state of the target as teaching data to update the parameters so as to improve the degree of agreement by the judgment unit.

Images