Sensor system

Abstract

The present invention provides a sensor system capable of efficiently extracting a feature amount from a sensor detection value. An analysis unit (48) of a sensor system (1) performs at least one of using at least one value among sensor detection values as a reference value to generate an analysis value on the basis of the reference value, generating the analysis value by statistically analyzing the same type or a plurality of types of feature amounts of the sensor detection values, and generating the analysis value by referring to data indicating a relationship between the sensor detection value and a state of a structure. Disturbance influence included in the analysis value is less than disturbance influence included in the sensor detection value.

Description

Sensor System

[0001] The present invention relates to a sensor system.

[0002] There are devices for inspecting infrastructure structures. Examples of infrastructure structures include roads, railways, bridges, retaining walls, and slopes. Infrastructure structures are also called infrastructure. Inspection of infrastructure structures is performed by placing sensors on the infrastructure structures. Patent Document 1 discloses a composite sensor module used for inspecting infrastructure structures.

[0003] Japanese Patent Application Laid-Open No. 2020-159835

[0004] A social infrastructure structure is called a structure. A value that indicates the state of a structure is called a feature. Examples of feature are the physical quantity of a structure and the behavior of a structure. A value detected by a sensor is called a sensor detection value. Assume that the sensor is placed in a structure. The sensor detection value may be affected by the environment of the structure. Examples of factors included in the environment that affect the sensor detection value are temperature, humidity, moisture, light, and wind.

[0005] The influence from the environment is called a disturbance. The influence of the disturbance contained in the sensor detection value is called disturbance noise. The sensor detection value may include a value indicating a feature amount and disturbance noise.

[0006] So far, no sensor system has been proposed that can efficiently remove disturbance noise from sensor detection values.

[0007] Sensors to be placed in structures include flexible sensors (low-rigidity sensors that are resistant to bending, stretching, twisting, etc.) and rigid sensors that are difficult to bend. It is preferable that sensors to be placed in structures be flexible. Flexible sensors are easily placed on surfaces of various shapes, such as curved surfaces. Flexible sensors are also less likely to affect the physical quantities and behavior of the structure. Flexible sensors are also more likely to detect small changes in feature quantities.

[0008] However, it is difficult to impart a structure that makes a flexible sensor less susceptible to disturbances while maintaining the sensor's flexibility, and therefore flexible sensors are more susceptible to disturbances than rigid sensors.

[0009] An object of the present disclosure is to provide a sensor system capable of efficiently extracting feature quantities from sensor detection values, particularly from sensor detection values ​​detected by a flexible and bendable sensor.

[0010] The sensor system includes a plurality of flexible sheet-like sensors that are placed on a structure and detect values ​​that indicate the state of the structure, and an analysis unit that analyzes the sensor detection values ​​when the values ​​detected by the sensors are taken as sensor detection values, and when the analysis unit takes as an analytic value a value generated by analyzing the sensor detection values, the analysis unit performs at least one of the following: using at least one value among the sensor detection values ​​as a reference value and generating the analytic value based on the reference value; generating the analytic value by performing statistical analysis of the same or multiple types of feature quantities among the sensor detection values; and generating the analytic value by referring to data that indicates the relationship between the sensor detection values ​​and the state of the structure, and the analytic value contains less external disturbance influence than the sensor detection values.

[0011] According to the present invention, it is possible to provide a sensor system that can efficiently extract feature quantities from sensor detection values.

[0012] 1 is a diagram showing the functional configuration of the sensor system of this embodiment. FIG. 1 is a diagram showing how sensors are arranged in a structure. FIG. 2 is a diagram showing a processing unit 3 of a sheet-type sensor system. FIG. 2 is a block diagram showing the configuration of the sensor system. FIG. 3 is a diagram showing the waveform of the applied voltage in the impedance sensor unit. FIG. 4 is a diagram showing the arrangement of concrete electrodes in the self-potential sensor unit. FIG. 5 is a diagram showing the arrangement of strain sensors in the strain sensor unit. FIG. 6 is a diagram showing the arrangement of thermistors in the thermistor unit. FIG. 7 is a graph showing changes in battery voltage of a thin lithium battery due to measurement. FIG. 8 is a table showing parts that are powered on at each timing. FIG. 9 is a diagram showing the operation flow of the sensor system. FIG. 10 is a graph showing the temperature and amount of water vapor in the environment. FIG. 11 is a diagram showing sensor detection values ​​in which the influence of disturbances is not reduced. FIG. 12 is a diagram showing sensor detection values ​​in which the influence of disturbances is reduced. FIG. 13 is a diagram showing the entire structure. FIG. 14 is an enlarged view of the structure. FIG. 15 is a diagram showing sensor detection values ​​in which the influence of disturbances is not reduced. FIG. 16 is a diagram showing sensor detection values ​​in which the influence of disturbances is reduced. FIG. 17 is a graph showing sensor detection values ​​of a strain sensor. FIG. 18 is a graph showing intensity after principal component analysis. FIG. 19 is a diagram showing the strength of correlation between sensor detection values. FIG. 20 is a graph showing sensor values ​​without compensation for changes in temperature and humidity over time. FIG. 1 is a graph showing sensor values ​​compensated for changes in temperature and humidity over time; FIG. 2 is a diagram showing an overview of multi-channel signal processing; FIG. 3 is a diagram showing an overview of sensor data acquisition by intermittent operation; FIG. 4 is a diagram showing an overview of multi-channel signal processing; FIG. 5 is a diagram showing an overview of sensor data acquisition by intermittent operation; FIG. 6 is a diagram showing raw data from a strain sensor; FIG. 7 is a diagram showing an example of an installation location for placing sensors outdoors; FIG. 8 is a diagram showing the results of long-term monitoring; FIG. 9 is a diagram showing the relationship between correlations between sensor data and precipitation labeling; FIG. 10 is a diagram showing daily precipitation; and FIG. 11 is a diagram showing impedance after compensation.

[0013] (Sensor System) A sensor system 1 of this embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing the functional configuration of the sensor system 1. The sensor system 1 includes a sensor unit 2 and a processing unit 3. The sensor unit 2 is a part that includes a sensor 13 disposed in a structure. The processing unit 3 is a part that drives the sensor 13 included in the sensor unit 2.

[0014] (Sensor Section) The sensor section 2 includes various sensors 13. In the example shown in Figure 1, the sensor 13 includes a temperature and humidity sensor section 14, a concrete electrode pair, a concrete rebar connection section 18, a concrete electrode 19, a strain sensor, and a thermistor 22.

[0015] The concrete electrode pairs include a first concrete electrode pair 15, a second concrete electrode pair 16, and a third concrete electrode pair 17. The strain sensors include a first strain sensor 20 and a second strain sensor 21.

[0016] The temperature and humidity sensor unit 14 is a part that measures the environmental temperature and humidity. The concrete electrode pair is an electrode pair for measuring impedance. The concrete rebar connection part 18 and the concrete electrode 19 are components for measuring natural potential. The strain sensor is a sensor for measuring strain. The thermistor 22 is a component for measuring temperature.

[0017] The sensor 13 may be of various types, such as a strain sensor, vibration sensor, pressure sensor, tilt sensor, acceleration sensor, temperature sensor, humidity sensor, water level sensor, rainfall sensor, ion sensor, corrosion sensor, moisture sensor, potential sensor, and impedance sensor. It is more preferable that the entire sensor system 1 is installed in a position adjacent to or in the same space as the sensor 13. Details of the sensor 13 will be described later.

[0018] (Processing Unit) The processing unit 3 includes a sensor driving unit 5, a battery power supply unit 6, an intermittent operation control unit 7, a power supply control unit 8, a central control unit 4, and a communication unit 10. The central control unit 4 includes a CPU 9. The sensor driving unit 5 is a unit that drives the sensors included in the sensor unit 2. The battery power supply unit 6 is a unit that supplies power to each unit of the sensor system 1. The intermittent operation control unit 7 is a unit that temporarily suspends the supply of power to circuits included in the sensor system 1. The purpose of intermittent operation is to reduce power consumption in the sensor system 1. Intermittent operation will be explained later.

[0019] The power supply control unit 8 is a unit that controls the power supplies of the sensor unit 2, communication unit 10, etc. The central control unit 4 is a unit that controls the entire sensor system 1. The CPU 9 is a component that performs various calculations in the sensor system 1. An example of the central control unit 4 including the CPU 9 is a microcomputer with a maximum operating frequency of 80 MHz and 246 Kbytes of flash memory, or a microcomputer board that has such a microcomputer installed. The communication unit 10 is a unit that transmits values ​​detected by the sensor from the sensor system 1 to a server or the like via a receiver located in a different location from the sensor system 1 or in a remote location. Examples of data communication by the communication unit 10 are a maximum of 100 bytes per hour and a maximum of 1 Kbyte per day.

[0020] (Sheet-type sensor) A sheet-shaped sensor is called a sheet-type sensor. The sensor 13 is a sheet-type sensor, and the thickness of its substrate and sealing material is preferably approximately 1 μm or more and 1 mm or less. Sheet-type sensors are flexible. Sheet-type sensors are preferably flexible and durable, with a bending radius of approximately 1 mm or more and 10 cm or less. Sheet-type sensors are stretchable, and the stretch rate (the ratio of deformation to the initial length; +100% strain = 2x elongation) is preferably 5% to 10% strain, which occurs when rebar or concrete is significantly deformed, and more preferably 10% to 50% strain, which occurs when large deformations such as cracks occur. Examples of methods for improving flexibility include sensors with a horseshoe shape (two-dimensional or three-dimensional spring shape) or composite conductors or semiconductors formed on a flexible substrate. Organic materials such as polymer resins, elastomers, and cellulose paper are used for the flexible substrate. A composite material system is a material system in which conductors or semiconductors (such as metal nanowires, metal nanoparticles, metal microparticles, carbon nanoparticles, carbon microparticles, and π-conjugated compounds) are dispersed or layered in an organic material. Examples of the sensor 13 include sensors that exhibit resistance changes, impedance changes, dielectric constant changes, electromotive force changes (changes in voltage output or current output), transistor characteristics changes, capacitor characteristics changes, ion-sensitive membrane potential changes, and potential changes as responses to the surrounding environment, physical quantities, and behavior of the structure. Sensors 13 also include sensors that directly probe (detect) the surrounding environment, physical quantities, and behavior of the structure using two-terminal or four-terminal techniques. The sensor 13 is disposed on the concrete surface 106 of the structure 100. The structure 100 and the concrete surface 106 are shown in FIG. 2A and other figures. The sensor 13 may also be disposed in the soil, slope, space, water, or water surface adjacent to the structure.

[0021] If the sensor does not have flexibility, the placement of the sensor may affect changes in the structure. In the sensor system 1 of this embodiment, the sensor 13 has flexibility, so the effect of the placement of the sensor 13 on changes in the structure 100 is suppressed.

[0022] (Structure) An example in which the sensor 13 is disposed in a structure 100 will be described. Fig. 2A is a diagram showing the sensor 13 disposed in the structure 100. Fig. 2B is a diagram showing the processing unit 3 in a flexible sheet-type sensor system that is particularly durable against bending. The processing unit 3 shown in Fig. 2B has a size of 10 cm x 25 cm x 60 µm and is flexible. The sensor 13 is disposed on a concrete surface 106 of the structure 100. The structure 100 shown in Fig. 2A is a bridge. Fig. 2A shows the bridge as viewed from below. Fig. 2A mainly shows the deck slab 102 and the skeleton 104 of the bridge.

[0023] 2A , the temperature and humidity sensor unit 14, the first concrete electrode pair 15, the second concrete electrode pair 16, the first strain sensor 20, and the second strain sensor 21 are arranged on the concrete surface 106 of the underside of the deck slab 102. The processing unit 3 is arranged on the concrete surface 106 of the skeleton 104. Two temperature and humidity sensor units 14 are arranged.

[0024] In the sensor system 1, the sensor 13 and the processing unit 3 are arranged in the structure 100. An example of the configuration of the processing unit 3 is shown in FIG. 2B. Values ​​detected by the sensor 13 are transmitted from the sensor system 1 to a server (not shown). Transmission can be performed, for example, by Sigfox. Detection by the sensor 13 can be performed intermittently at, for example, 1 data / H. Data transmission and intermittent operation will be described later.

[0025] Each unit of the sensor system 1 will now be described in more detail. Fig. 3 is a block diagram showing the configuration of the sensor system 1. (Sensor Driving Unit) The sensor driving unit 5 includes an impedance sensor unit 25, an ADC unit 45, and a multiplexer 46. The sensor driving unit 5 also includes a part of the spontaneous potential sensor unit 26, a part of the strain sensor unit 27, and a part of the thermistor unit 28.

[0026] The impedance sensor unit 25 applies a sinusoidal voltage to the concrete electrode pair to measure the impedance of the structure 100. The ADC unit 45 converts analog signals into digital signals. The resolution of the ADC unit 45 is preferably between 12 and 24 bits. The multiplexer 46 combines multiple signals into a single signal. The multiplexer 46 combines the signal from the spontaneous potential sensor unit 26, the signal from the strain sensor unit 27, and the signal from the thermistor unit 28 into a single signal. The signal combined by the multiplexer 46 is input to the ADC unit 45.

[0027] The self-potential sensor unit 26 is a unit that measures the voltage generated in the structure 100. The strain sensor unit 27 is a unit that measures expansion / contraction and cracks of the structure 100. The thermistor unit 28 is a unit that measures the temperature of the structure 100.

[0028] (Impedance Sensor Unit) The impedance sensor unit 25 includes a sine wave generator 30, an impedance calculator 31, and a multiplexer 32. The sine wave generator 30 generates a sine wave to be applied to the concrete electrode pair, and the impedance calculator 31 examines the current from the concrete electrode pair and calculates the impedance. The multiplexer 32 combines signals from the first concrete electrode pair 15, the second concrete electrode pair 16, and the third concrete electrode pair 17 into one. The sine wave generator 30, the impedance calculator 31, and the multiplexer 32 are connected in this order.

[0029] The detection process in the impedance sensor unit 25 will be described with reference to Fig. 4. Fig. 4 is a diagram showing the waveform of the voltage applied to the impedance sensor unit 25. Fig. 4 shows only one first concrete electrode pair 15. Therefore, Fig. 4 does not show the multiplexer 32.

[0030] The two electrodes included in the first concrete electrode pair 15 are each placed on the concrete surface of the structure 100. A voltage or current of a sine wave 35 generated by a sine wave generating unit 30 is applied to the first concrete electrode pair 15. An impedance calculating unit 31 checks the current or voltage based on a waveform 36 from the first concrete electrode pair 15. In this manner, the impedance is measured.

[0031] 4 may be configured as a single component. That is, the sine wave generating unit 30 and the impedance calculating unit 31 may be configured as a single component. An example of the single component is an impedance converter IC 38 that can measure complex impedances of 1 kΩ to 10 MΩ at a specific frequency between 1 kHz and 100 kHz and that includes a 12-bit ADC or the like.

[0032] (Self-Potential Sensor Unit) The self-potential sensor unit 26 includes a relay 40, a concrete rebar connection unit 18, and a concrete electrode 19. The concrete rebar connection unit 18 and the concrete electrode 19 are included in the sensor unit 2. The relay 40 is included in the processing unit 3.

[0033] The concrete rebar connection portion 18 and the concrete electrode 19 are disposed in the structure 100. The concrete rebar connection portion 18 and the concrete electrode 19 are connected to a relay 40.

[0034] The relay 40 is connected to a multiplexer 46 and an ADC unit 45 in this order.

[0035] The measurement process in the natural potential sensor unit 26 will be described with reference to Fig. 5. Fig. 5 is a diagram showing the arrangement of the concrete electrodes 19 in the natural potential sensor unit 26. Fig. 5 shows only the natural potential sensor unit 26. Therefore, Fig. 5 does not show the multiplexer 32.

[0036] 5, the concrete rebar connection portion 18 is connected to a rebar 108 of the structure 100. The concrete electrode 19 is placed on the concrete surface 106 of the structure 100.

[0037] This makes it possible to measure the voltage generated in the concrete of the structure 100. An example of the voltage measurement range is from −1.5 V to +1.5 V. The relay 40 is disposed for the purpose of establishing an insulating state to protect the charge generated between the concrete rebar connection part 18 and the concrete electrode 19.

[0038] (Strain Sensor Unit) The strain sensor unit 27 includes a first strain sensor 20, a second strain sensor 21, and a strain sensor driving unit 41. The first strain sensor 20 and the second strain sensor 21 are included in the sensor unit 2. The strain sensor driving unit 41 is included in the processing unit 3.

[0039] The first strain sensor 20 and the second strain sensor 21 are disposed on the structure 100. The first strain sensor 20 and the second strain sensor 21 are connected to a strain sensor driving unit 41.

[0040] The strain sensor driving section 41 is connected to a multiplexer 46 and an ADC section 45 in this order.

[0041] The measurement in the strain sensor unit 27 will be described with reference to Fig. 6. Fig. 6 is a diagram showing the arrangement of the first strain sensors 20 in the strain sensor unit 27. Fig. 6 shows only the strain sensor unit 27. Therefore, Fig. 6 does not show the multiplexer 32.

[0042] 6, the first strain sensor 20 is placed on the concrete surface 106 of the structure 100. The strain sensor is configured so that the resistance between predetermined portions on the concrete surface 106 can be measured.

[0043] The strain sensor driver 41 applies a voltage or current to the first strain sensor 20 and reads the current or voltage value corresponding to the change in resistance of the strain sensor, or reads the voltage value divided by a bridge circuit or the like. This allows the strain sensor unit 27 to measure the resistance value that changes due to factors such as the expansion and contraction of the concrete in the structure 100. The strain sensor unit 27 measures the amount of strain and crack width from the change in resistance value. An example of the resistance value to be measured is 10 Ω or more and 100 KΩ or less.

[0044] (Thermistor Section) The thermistor section 28 includes a thermistor 22 and a thermistor driving section 42. The thermistor 22 is included in the sensor section 2. The thermistor driving section 42 is included in the processing section 3.

[0045] The thermistor 22 is disposed in the structure 100. The thermistor 22 is connected to a thermistor driver 42.

[0046] The thermistor driving section 42 is connected to a multiplexer 46 and an ADC section 45 in this order.

[0047] The measurement in the thermistor section 28 will be described with reference to Fig. 7. Fig. 7 is a diagram showing the arrangement of the thermistors 22 in the thermistor section 28. Fig. 7 shows only the thermistor section 28. Therefore, Fig. 7 does not show the multiplexer 32.

[0048] As shown in Figure 7, the thermistor 22 is placed on the concrete surface 106 of the structure 100. The thermistor driver 42 applies a voltage to the thermistor 22 and checks the divided voltage. This allows the thermistor unit 28 to measure a resistance value that changes with temperature. The thermistor unit 28 measures the temperature from the change in resistance value.

[0049] (Battery power supply unit) The battery power supply unit 6 will now be described. As shown in Fig. 3, the battery power supply unit 6 includes a first battery power supply unit 50 and a second battery power supply unit 51. The first battery power supply unit 50 is a battery power supply unit that includes a battery that needs to be recharged or replaced when it runs out. The second battery power supply unit 51 is a battery power supply unit that includes a battery that does not need to be replaced, such as a solar cell.

[0050] The first battery power supply unit 50 and the second battery power supply unit 51 are connected to the intermittent operation control unit 7 .

[0051] (First Battery Power Supply Unit) The first battery power supply unit 50 includes a lithium battery 53 and a charge control IC 54. The charge control IC 54 is connected to the lithium battery 53. The lithium battery 53 is connected to the intermittent operation control unit 7. An example of the lithium battery 53 is a lithium battery pack. Power is supplied from the lithium battery 53 to the intermittent operation control unit 7.

[0052] When the lithium ion battery 53 is used up, the lithium battery pack is removed from the sensor system 1. Then, after the lithium battery pack is charged, the lithium battery pack is attached to the sensor system 1.

[0053] Alternatively, when the lithium ion battery 53 is used up, the used-up lithium battery pack is removed from the sensor system 1 and another charged lithium battery pack is attached to the sensor system 1. The first battery power supply unit 50 is a power supply that includes a battery that needs to be recharged when used up.

[0054] (Second Battery Power Supply Unit) The second battery power supply unit 51 includes a thin (small-capacity) lithium battery 56, a low-power charging control IC 57, a step-up DC-DC converter 58, and a solar cell 59. The thin (small-capacity) lithium battery 56, the low-power charging control IC 57, the step-up DC-DC converter 58, and the solar cell 59 are connected in this order. The thin (small-capacity) lithium battery 56 is also connected to the intermittent operation control unit 7.

[0055] The second battery power supply unit 51 includes a solar cell 59. Unlike the first battery power supply unit 50, the second battery power supply unit 51 does not require battery replacement. An example of the solar cell 59 is an amorphous silicon solar cell. The thin lithium battery 56 may also be a sheet-type lithium battery.

[0056] The voltage of the solar cell 59 is between 4.5 V and 6 V. On the other hand, the operating condition of the small power charging control IC 57 is 5 V. Therefore, the output voltage of the solar cell 59 is boosted by the boost DC-DC converter 58.

[0057] (Intermittent operation control unit) The intermittent operation control unit 7 will now be described. The intermittent operation control unit 7 includes an FET (power switch) 61, a flip-flop 62, a voltage determination unit 63, a timer IC 64, and a button battery 65. The flip-flop 62, the voltage determination unit 63, the timer IC 64, and the button battery 65 are connected to the FET (power switch) 61 in this order. The FET (power switch) 61 is also connected to the lithium battery 53 of the first battery power supply unit 50 and the thin (small capacity) lithium battery 56 of the second battery power supply unit 51. The FET (power switch) 61 is also connected to the CPU 9 and the power supply control unit 8.

[0058] In conventional intermittent operation, the CPU and its power supply components (such as the regulator) remain powered while waiting for a CPU interrupt and reset. The intermittent operation control unit 7 of this embodiment operates the sensor system 1 in a fully intermittent manner. Fully intermittent operation is an operation in which the power supply to the entire circuit, including the CPU 9, is completely stopped until the timer is started. Fully intermittent operation achieves maximum power savings.

[0059] The following describes a case where the sensor system 1 operates using a thin (small-capacity) lithium battery 56 as a power source. In this case, the intermittent operation control unit 7 starts various measurements after the thin (small-capacity) lithium battery 56 has been charged to a predetermined amount.

[0060] The voltage of the thin (small-capacity) lithium battery 56 drops during measurement. Therefore, the intermittent operation control unit 7 charges the thin (small-capacity) lithium battery 56 to account for the voltage drop caused by the measurement before starting measurement. This makes it possible to avoid measurement failures caused by starting measurement at the lowest operating voltage.

[0061] The fluctuation of the battery voltage due to measurement and charging will be described with reference to Fig. 8. Fig. 8 is a diagram showing the change in the battery voltage due to measurement and charging of a thin (small capacity) lithium battery 56.

[0062] 8 indicates the start of measurement 122. Point 123 indicates the end of sensor measurement 123. Point 124 indicates the end of Sigfox communication 124. Arrow 126 indicates one measurement 126. One measurement 126 is completed through the flow of measurement start 122, sensor measurement end 123, and Sigfox communication end 124.

[0063] (Intermittent Operation) The sensor system 1 of this embodiment operates completely intermittently. As indicated by arrow 127 in Fig. 8, a period during which measurement 126 is repeated is called a measurement period 127. As indicated by arrow 128, a period during which measurement 128 is paused is called a power-off period 128. In completely intermittent operation, the measurement period 127 and the power-off period 128 are repeated.

[0064] (Charging Period) The power supply voltage of the thin (small capacity) lithium battery 56 drops from the battery voltage at the start 122 of the measurement to the battery voltage at the end 124 of the Sigfox communication in one measurement 126 .

[0065] Figure 8 shows a case where the voltage used to determine whether or not to start measurement 122 is 3.8V. Line 120 in Figure 8 indicates a battery voltage of 3.8V. Intermittent operation control unit 7 charges thin (small-capacity) lithium battery 56 so that the battery voltage at the time of measurement start 122 does not fall below 3.8V, and then starts measurement 122. Arrow 129 indicates charging period 129, which is the period during which thin (small-capacity) lithium battery 56 is being charged.

[0066] (Power Supply Control Unit) The power supply control unit 8 will be described with reference to Fig. 3. The power supply control unit 8 includes an FET (power switch) 70 and an I / O expansion 71. The I / O expansion 71 is connected to the FET (power switch) 70. The FET (power switch) 70 is also connected to the FET (power switch) 61 of the intermittent operation control unit 7. The FET (power switch) 70 is also connected to the CPU 9, the sensor unit 2, and the communication unit 10.

[0067] In the sensor system 1, the power supply for each component is separate. The power supply control unit 8 supplies power only to the components required for measurement and communication. This reduces the power consumption of the sensor system 1. This reduces the consumption of the lithium battery 53 and the thin (small-capacity) lithium battery 56. As a result, long-term measurement becomes possible.

[0068] (Power Supply Control) Referring to Fig. 9, the parts that are powered on at each timing will be described. Fig. 9 is a diagram showing the parts that are powered on at each timing. Each vertical item in the table of Fig. 9 indicates each timing. Each horizontal item in the table of Fig. 9 indicates each part of the sensor system 1. Check marks in Fig. 9 indicate parts that are powered on.

[0069] As shown in Figure 9, different parts are powered on at each timing. At the end of the period, only the timer IC is powered on. This allows for maximum power savings during fully intermittent operation.

[0070] (Communication Unit) The communication unit 10 will be described with reference to Fig. 3. The communication unit 10 includes a first communication unit 75, a second communication unit 76, and a third communication unit 77. The first communication unit 75 is configured as a Wi-Fi communication unit 80. The second communication unit 76 is configured as a Sigfox communication unit 81.

[0071] The third communication unit 77 functions as an LTE (registered trademark)-M communication unit. The third communication unit 77 includes an LTE (registered trademark)-M SIM 83, a SIM conversion IC 84, and a communication CPU 85. The LTE (registered trademark)-M SIM 83, the SIM conversion IC 84, and the communication CPU 85 are connected in this order.

[0072] The first communication unit 75, the second communication unit 76, and the third communication unit 77 are all units that transmit data from the sensor system 1 to a server or the like via wireless communication. The first communication unit 75 transmits data to a server or the like via a router, a base station, or the like. The second communication unit 76 transmits data to a server or the like via a Sigfox gateway. The third communication unit 77 transmits data to a server or the like via a base station.

[0073] The user of the sensor system 1 collects data from the server via the Internet and can view the data using a personal computer, a smartphone, or the like.

[0074] An example of the Wi-Fi communication unit 80 of the first communication unit 75 is a module compatible with Wi-Fi 802.11 b / g / n, or a microcomputer board equipped with such a module. An example of the Sigfox communication unit 81 of the second communication unit 76 is a module compatible with Sigfox Zone 3, or a breakout board equipped with such a module. An example of the SIM conversion IC 84 of the third communication unit 77 is a module compatible with LTE (registered trademark) Cat. M1, or a communication board equipped with such a module. An example of the communication CPU 85 of the third communication unit 77 is a microcomputer equipped with an 8-bit processor (maximum clock frequency 20 MHz, SRAM 6 KB), or a microcomputer board equipped with such a microcomputer.

[0075] The first communication unit 75, the second communication unit 76, and the third communication unit 77 of the communication unit 10 communicate with the inside of the CPU 9 via serial communication.

[0076] (Analysis Unit) The processing unit 3 includes an analysis unit 48. The analysis unit 48 is a part that analyzes the values ​​detected by the sensor 13 after the fact or in real time. The analysis unit 48 may be arranged to process data stored on a server or in the cloud, or may be arranged within the sensor system 1 that is partially or entirely composed of sheet-shaped or printable electronic circuits.

[0077] The processing unit 3 may include an analog signal processing unit 11 and a data storage unit 12. The analog signal processing unit 11 is a unit that processes analog signals within the processing unit 3. The data storage unit 12 is a unit that stores the results detected by the sensor 13, etc.

[0078] The processing unit 3 may include at least one of an output unit 150, a light-emitting unit 151, a wireless power supply unit 152, and an independent power supply unit 153. The output unit 150 is a unit that outputs the results detected by the sensor 13 and the state of the structure, etc. The output can be performed by displaying on a display device, etc. The light-emitting unit 151 is a unit that notifies the outside of an abnormality by emitting light. The light-emitting unit 151 emits light when the results detected by the sensor 13 are abnormal, when the state of the structure is abnormal, etc.

[0079] The wireless power supply unit 152 is a unit that receives power wirelessly from outside the sensor system 1. The supplied power is supplied to the lithium battery 53 or the like. The independent power supply unit 153 is a unit that receives power supplied by an independent power supply from outside the sensor system 1. The supplied power is supplied to the lithium battery 53 or the like. Power supplied by an independent power supply refers to power generated by a source other than a commercial power source, such as an engine generator, wind power generation, vibration power generation, or moisture power generation. This power may be stored in a storage battery, a lithium ion battery, a capacitor, a condenser, or the like until an independent power supply is achieved. The wireless power supply unit 152 and the independent power supply unit 153 may be used in place of the second battery power supply unit 51.

[0080] Each component included in the sensor system 1 may be configured, for example, as a sheet-shaped or printable electronic circuit. It is preferable that the thickness of the substrate or sealing material of the sheet-shaped or printable electronic circuit be approximately 1 μm or more and 1 mm or less. It is also preferable that the substrate or sealing material have flexibility, such as bending and stretching, to reduce the impact on the physical quantities and behavior of the structure. An example of a method for improving flexibility is to integrate and mount elements on a circuit board formed with a horseshoe-shaped (two-dimensional or three-dimensional spring) or composite-based conductor or semiconductor on a flexible substrate. Organic materials such as polymer resins, elastomers, and cellulose paper are used as flexible substrates. A composite material system is a material system in which conductors or semiconductors (e.g., metal nanowires, metal nanoparticles, metal microparticles, carbon-based nanoparticles, carbon-based microparticles, π-conjugated compounds, etc.) are dispersed or layered on an organic material.

[0081] (Operation Flow of Sensor System) The operation flow of the sensor system 1 will be described with reference to Fig. 10. Fig. 10 is a diagram showing the operation flow of the sensor system 1. In Fig. 10 and the following description, S1 means step 1. The same applies to S2 and the following steps.

[0082] At S1, the timer IC 64 outputs a signal after a certain time has elapsed. At S2, the voltage determination unit 63 passes the signal as is if the rechargeable battery voltage is higher than a predetermined voltage. At S3, when the signal is input, the flip-flop 62 turns its output ON and maintains the ON state. At S4, the FET (power switch) 61 turns ON, and the rechargeable battery voltage is sent to the CPU 9.

[0083] Steps S5 to S12 are executed under the control of the CPU 9. In S5, the power supply control unit 8 controls the power supplies of the sensor 13, the communication unit 10, and the like depending on the components being used. In S6, the impedance sensor unit 25 acquires a measured impedance value. In S7, the natural potential sensor unit 26 acquires a natural potential voltage. In S8, the strain sensor unit 27 acquires a voltage representing the strain measurement result. In S9, the first communication unit 75 (Wi-Fi communication unit 80) transmits the measurement result via Wi-Fi. In S10, the second communication unit 76 (Sigfox communication unit 81) transmits the measurement result via Sigfox. In S11, the third communication unit (LTE (registered trademark)-M communication unit) 77 transmits the measurement result via LTE (registered trademark)-M. In S12, the flip-flop 62 is set to the OFF state. This turns off the power supply of the CPU 9 itself.

[0084] The series of operations from S1 to S12 as described above is completed. Note that during the flow from S6 to S8, the thermistor unit 28 and the temperature / humidity sensor unit 14 related to the sensor unit 2 and the sensor driving unit 5 may be selectively operated. At least one of the sensor unit 2 and the sensor driving unit 5 is selectively operated at least once during the flow. At least one of the communication units S9, S10, and S11 is selectively operated.

[0085] (Sensor Detected Values) The sensor detected values ​​and the processing thereof will be described. As described above, the sensor detected values ​​may contain disturbance noise. In the sensor system 1 of this embodiment, when the analysis unit 48 sets a value generated by analyzing the sensor detected values ​​as an analytic value, the analysis unit 48 performs at least one of the following: 1. sets at least one value of the sensor detected values ​​as a reference value and generates an analytic value based on the reference value; 2. generates an analytic value by statistically analyzing the same or multiple types of feature quantities of the sensor detected values; and 3. generates an analytic value by referring to data indicating the relationship between the sensor detected values ​​and the state of the structure, and the analytic value contains less disturbance influence than the sensor detected values.

[0086] The analysis unit 48 may generate the analysis value by at least one of determining the difference between the reference value and the sensor detection value other than the reference value and determining the ratio of the sensor detection value other than the reference value to the reference value.

[0087] The reference value may be generated based on a sensor detection value acquired by a sensor 13 placed either at a position in a healthy state of the structure 100 or at a position in a healthy state that simulates the structure 100.

[0088] When the analysis unit 48 statistically analyzes the same or multiple types of feature quantities among the sensor detection values, the analysis unit 48 may generate an analysis value by extracting the feature quantities of the analysis value by principal component analysis.

[0089] When the analysis unit 48 refers to the status data indicating the relationship between the sensor detection value and the status of the structure, the analysis unit 48 may generate the analysis value by referring to the analysis data or analytical data of at least one of the inspection and the image.

[0090] The sensor system 1 may output at least one of the feature amount and the health status as a final output.

[0091] The sensor system 1 may take at least one of feature data and status data as input, process the input data using time series data analysis technology to generate a predictive model in real time, and output at least one of the predicted feature data and health status as a final output.

[0092] The above configuration enables efficient derivation of true values ​​by correcting variations in sensor detection values ​​in consideration of correlation with changes in the external environment. An example will be described in which the disturbances are temperature and humidity.

[0093] (Influence of Temperature and Humidity) An example will be described in which the influence of temperature and humidity as disturbances is reduced by calculating the difference between sensor detection values. The sensor 13 is a carbon sensor. The carbon sensor is a composite material formed by printing a paste in which carbon particles and an organic material are dispersed. The sensor 13 is placed on a bridge. FIG. 11A is a graph showing the temperature and amount of water vapor in the environment of the sensor 13. The X-axis of the graph 200 represents the number of days. The Y1-axis of the graph 200 represents the temperature. The Y2-axis of the graph 200 represents the amount of water vapor. As shown in the graph 200, the environment fluctuates over the course of a year or from year to year. The temperature fluctuates between -5°C and 15°C. The humidity fluctuates between 50% RH and 100% RH. The humidity can be calculated from the temperature and amount of water vapor.

[0094] Fig. 11B is a diagram showing the sensor detection values ​​and the like in which the influence of disturbances has not been reduced. Graph 201 in Fig. 11B is a graph showing the sensor detection values ​​measured by sensor 13. The X-axis of graph 201 represents the number of days. The Y-axis of graph 201 represents the resistance value. Graph 201 shows the detected sensor detection values ​​as they are. In other words, graph 201 shows the sensor detection values ​​in which the influence of disturbances has not been reduced.

[0095] Graph 201 shows the sensor detection values ​​of eight sensors 13. CH1 to CH8 show the respective sensor detection values. CH1 to CH4 show the sensor detection values ​​of sensors 13 placed in parts of the bridge where there are cracks. CH5 to CH7 show the sensor detection values ​​of sensors 13 placed in parts of the bridge where there are no cracks. CH8 shows the sensor detection value of a sensor 13 placed for reference. The reference sensor 13 is placed in the air.

[0096] Graph 202 in Figure 11B is a graph showing values ​​measured by a conventional strain gauge. The X-axis of graph 202 shows the number of days. The Y-axis of graph 202 shows strain. CH000 to CH003 show the gauge values, respectively. CH002 and CH000 show the values ​​of the gauge placed in a part of the bridge where there is a crack. CH001 and CH003 show the values ​​of the gauge placed in a part of the bridge where there is no crack.

[0097] As shown in graph 202, the gauge values ​​differ between the gauge placed in the cracked portion and the gauge placed in the crack-free portion. In contrast, as shown in graph 201, the sensor detection value of sensor 13 does not clearly differ between the sensor placed in the cracked portion and the sensor placed in the crack-free portion.

[0098] The reason why the sensor detection values ​​of the sensor 13 in the area with a crack and the area without a crack do not clearly differ is because the sensor detection values ​​are affected by external disturbances, i.e., the changes in the sensor detection values ​​are affected by the temperature and humidity of the environment.

[0099] 11C is a graph showing the sensor detection value etc. after the influence of the disturbance has been reduced. Graph 203 in FIG. 11C is a graph showing the sensor detection value measured by sensor 13 after the influence of the disturbance has been reduced.

[0100] The X-axis of graph 203 represents the number of days. The Y-axis of graph 203 represents the resistance value. Graph 203 shows four differential signals for the sensor detection values ​​of sensor 13. Differential signal 1 represents the value of CH1 (cracks present) - CH5 (no cracks present). Differential signal 2 represents the value of CH3 (cracks present) - CH5 (no cracks present). Differential signal 3 represents the value of CH1 (cracks present) - CH6 (no cracks present). Differential signal 4 represents the value of CH3 (cracks present) - CH6 (no cracks present). As shown in graph 203, when the four differential signals are plotted, they nearly overlap.

[0101] Graph 204 in Fig. 11C is a graph in which CH003 is hidden from graph 202 in Fig. 11B. The changes shown by differential signals 1 to 4 in graph 203 closely resemble the changes shown by CH002 (crack present) and CH000 (crack present) in graph 204. This means that by finding the difference between the detected values ​​of the two sensors, it is possible to obtain results similar to those obtained by measurements using a conventional gauge. In other words, this shows that the effects of disturbances are reduced by finding the differential signal.

[0102] (Changes in Sensor Over Time) An example of reducing the effects of changes in the strain sensor over time will be described with reference to Fig. 12A to Fig. 12C. Fig. 12A is a diagram showing the entire structure 100. Fig. 12B is an enlarged view of the structure 100. Fig. 12C is a diagram showing sensor detection values ​​in which the effects of disturbances have not been reduced. Fig. 12D is a diagram showing sensor detection values ​​in which the effects of disturbances have been reduced.

[0103] 12A , the strain sensors are arranged on the piers of a bridge. The bridge is an example of a structure 100. The piers are an example of a skeleton 104.

[0104] Figure 12B is an enlarged view of a portion of the framed area 215 in Figure 12A. A framed area 217 in Figure 12B is an enlarged view of a portion of the framed area 216. As shown in the framed area 217, the first strain sensor 20 and the second strain sensor 21 are disposed on the concrete surface of the skeleton 104. In the example shown in the framed area 217, the first strain sensor 20 and the second strain sensor 21 are disposed across a crack 220.

[0105] The X-axis of graph 222 in Fig. 12C and graph 223 in Fig. 12D both represents the number of days. The Y-axis of graph 222 in Fig. 12C and graph 223 in Fig. 12D both represents the resistance value. Fig. 12C is a diagram showing the sensor detection values ​​as they are. In other words, Fig. 12C is a diagram showing the raw data of the sensor detection values. Plots 225 to 227 in Fig. 12C represent sensor detection values ​​detected by different strain sensors. Plots 225, 226, and 227 show different trends. This is because the strain sensors are affected by environmental factors such as temperature, humidity, and concrete strain, causing each of the sensor detection values ​​to exhibit free behavior.

[0106] Graph 223 in Fig. 12D shows the sensor detection value after canceling the influence of changes in the strain sensor over time. Plots 228 and 229 in graph 223 each show the difference from other sensor detection values. Plot 228 shows the value of plot 225 minus plot 226. Plot 229 shows the value of plot 227 minus plot 226.

[0107] By calculating the difference between the sensor detection values, it is possible to cancel out the change in the sensor over time.By calculating the difference between multiple plots, in other words, the sensor detection values ​​detected by multiple sensors, and other plots, it is possible to cancel out the change in the sensor over time in a long-term test, for example, over two years.

[0108] When calculating the difference, it is preferable to calculate the difference for one predetermined plot. The value of this predetermined plot is used as the reference value. In the case of graph 223, the reference value is the value of plot 226. It is possible to appropriately determine which sensor's detected value is used as the reference value when calculating the difference.

[0109] The sensor that detects the reference value is called the reference sensor. The method of reducing the influence of external factors by calculating the difference between the sensor value detected by the reference sensor is called the differential method.

[0110] As described above, the difference method cancels out the changes in the sensor over time. As a result, only the strain in the concrete of the bridge pier can be detected. In other words, it becomes possible to identify whether or not there are cracks on the concrete surface. In graph 223, plot 228 indicates the absence of cracks, and plot 229 indicates the presence of cracks.

[0111] The algorithm for canceling the influence of external factors on the sensor and the influence of changes in the sensor over time is not limited to the differential method, and for example, an algorithm such as principal component analysis can be used.

[0112] (Principal Component Analysis) Referring to Figures 13A to 13C, principal component analysis will be described as another example of constructing an algorithm. Figure 13A is a graph showing sensor detection values ​​of a strain sensor. In the graph of Figure 13A, the X axis represents the number of days. In the graph of Figure 13A, the Y axis represents strain.

[0113] CH1 to CH4 indicate the sensor detection values ​​of the sensors 13 placed in the cracked areas. CH5 to CH7 indicate the sensor detection values ​​of the sensors 13 placed in the crack-free areas. CH8 indicates the sensor detection value of the sensor 13 placed for reference. The reference sensor 13 is placed in the air.

[0114] The graph in Fig. 13A shows the sensor detection values ​​for CH1 to CH8 individually, and plots CHs with cracks, CHs without cracks, and CHs in the air without being classified.

[0115] 13B is a graph showing the intensities after principal component analysis, which shows the results of principal component analysis of CH2 and CH3 with cracks and CH5 and CH6 without cracks.

[0116] As shown in the graph of FIG. 13B, the intensities of CH2 and CH3 with cracks and CH5 and CH6 without cracks are significantly different after the principal component analysis.

[0117] 13C is a diagram showing the strength of correlation between sensors in a principal component extracted by principal component analysis. The numerical values ​​in the diagram indicate the strength of correlation. As shown in FIG. 13C, there is a positive correlation between CHs with cracks and CHs with cracks. There is a positive correlation between CHs without cracks and CHs without cracks. There is a negative correlation between CHs with cracks and CHs without cracks.

[0118] 13C, the sensor detection values ​​can be classified by examining the correlation between the sensor detection values. Furthermore, the sensor detection values ​​and the sensors can be classified according to the condition of the building in which the sensor is located.

[0119] In this way, by examining the strength of correlation between analytical values ​​based on sensor detection values, i.e., by performing principal component analysis, an algorithm can be constructed that suppresses the influence of external factors and enables accurate judgment of the condition of a structure.

[0120] The above algorithm suppresses the influence of external factors even in long-term indoor measurements. Figure 14A is a graph showing sensor values ​​without compensation for changes in temperature and humidity over time for the sensor. Figure 14B is a graph showing sensor values ​​with compensation for changes in temperature and humidity over time for the sensor. The X-axis in Figures 14A and 14B represents the number of days. The Y-axis in Figures 14A and 14B represents the resistance value. Figures 14A and 14B show sensor detection values ​​at the 1st, 105th, 230th, and 285th cycles.

[0121] 14A, without compensation for temperature and humidity changes over time, there is a maximum variation of 160 Ω between the four cycles, as shown by arrow 210. The difference between the four cycles, even when the difference is small, is 75 Ω, as shown by arrow 211.

[0122] As shown by arrow 213 in Fig. 14B, when compensation for changes in humidity and temperature over time is performed using the algorithm described above, the difference between the four cycles is approximately 10 Ω or less. In this way, suppression of external influences using the algorithm described above suppresses the influence of external factors even in long-term indoor measurements.

[0123] (Multi-Channel Signal Processing) Multi-channel signal processing will be described with reference to Fig. 15A. Fig. 15A is a diagram showing an overview of multi-channel signal processing. As shown in Fig. 15A, the sensor system 1 includes, as parts that perform multi-channel signal processing, an AFE (Analog Front End) circuit 250, a selector 1 to X 251, an Amp. (Amplifier) ​​252, an ADC (Analog Digital Converter) 253, and a CPU (Central Processing Unit) 254. The AFE circuit 250, the selector 1 to X 251, the Amp. 252, the ADC 253, and the CPU 254 are connected in this order.

[0124] The sensor output channels are designated CH0 to CHX. The sensor outputs of these channels are input to the AFE circuit 250. The AFE circuit 250 outputs the sensor values ​​to a selector 1 to X 251. The selector 1 to X 251 outputs the sensor values ​​to an Amp. 252. The sensor value that the selector 1 to X 251 outputs to an Amp. 252 is determined by a signal selection from a CPU 254. The Amp. 252 amplifies the sensor value. The Amp. 252 outputs the amplified sensor value to an ADC 253. The ADC 253 A / D converts the sensor value. The ADC 253 outputs the A / D converted sensor value to the CPU 254. The CPU 254 processes the sensor values ​​and outputs them as sensor data Data0 to DataX. The sensor data Data0 corresponds to the sensor output CH0. The same correspondence applies to other sensor data.

[0125] (Sensor Data Acquisition Through Intermittent Operation) An overview of sensor data acquisition through intermittent operation will be described with reference to FIG. 15B. FIG. 15B is a diagram showing an overview of sensor data acquisition through intermittent operation. One cycle 130 of intermittent operation includes a measurement period 127 and a power-off period 128. During the measurement period 127, the sensors corresponding to CH0 to CHX measure in order. When the measurement corresponding to CHX ends, the power-off period 128 begins. When the predetermined power-off period 128 has elapsed, the measurement period 127 of the next cycle 130 begins. The cycle 130 includes the battery voltage determination described with reference to FIG. 8.

[0126] An example of multi-channel signal processing will be described with reference to Figures 16A to 16C. Figure 16A is a diagram showing an overview of multi-channel signal processing. Figure 16B is a diagram showing an overview of sensor data acquisition by intermittent operation. Differences between Figures 16A and 16B and Figures 15A and 15B will be mainly described. In the example shown in Figures 16A to 16C, the number of channels is 8. In Figures 16A and 16B, CHX in Figures 15A and 15B becomes CH7.

[0127] As shown in Figure 16A, the strain sensor outputs CH0 to CH7. The AFE 250 converts changes in sensor resistance due to strain fluctuations into voltage changes. The selector 8to1 255 receives selection signals A0 to A2 from the CPU 254. The selector 8to1 255 outputs the sensor value to the Amp. 252 in accordance with the received selection signal. The sensor value is input to the CPU via the ADC 253. The CPU outputs Data0 to Data7 as raw strain sensor data.

[0128] Intermittent operation will be described with reference to Fig. 16B. The intermittent operation shown in Fig. 16B is an intermittent operation that occurs once per hour. The number of channels is eight. Measurements are made on CH0 to CH7 once per one-hour cycle 130. The measurement period 127 is several seconds.

[0129] Data processing will be described. Of the strain sensor raw data 254, Data 0 is the raw data of the reference sensor, and Data 1 to Data 7 are the raw data of the measurement target sensor.

[0130] Fig. 16C is a diagram showing strain sensor raw data. In the graph shown in Fig. 16C, the X axis represents the date. The Y1 axis represents the strain value. Y2 represents the temperature and humidity. Plots 281 to 283 represent the strain sensor raw data. Plot 285 represents humidity. Plot 286 represents temperature.

[0131] As plots 281 to 283 show, the strain sensor raw data values ​​are affected by humidity and temperature.

[0132] Therefore, to correct for variations in temperature and humidity, as well as variations in the sensors over time, the difference between the raw data of each target sensor and the raw data of the reference sensor is calculated. The difference between the raw data of the target sensor and the raw data of the reference sensor is called the distortion value.

[0133] Data1-Data0 is the CH1 strain value. Data2-Data0 is the CH2 strain value. Data3-Data0 is the CH2 strain value. By determining strain values ​​in this manner, sensor detection values ​​in which the influence of external factors is suppressed can be obtained.

[0134] The calculation for finding the difference, in other words, the introduction of the algorithm described above, may be performed on the CPU 254 side, the server side, or the like.

[0135] The multi-channel remote monitoring device can have a built-in solar cell. By incorporating a solar cell, battery replacement is not necessary. The multi-channel remote monitoring device simultaneously measures one or more types of physical quantities. Examples of physical quantities include temperature, humidity, moisture, strain, cracks, electric potential, impedance, dielectric constant, electromagnetism, light, ions, pressure, vibration, and magnetism.

[0136] The autonomous multi-channel measurement device (sensor system) may transmit measurement data wirelessly. Examples of wireless communication include Wi-Fi, Sigfox, and RFID. By distributing the autonomous multi-channel measurement devices at multiple points on a structure, it is possible to obtain a large amount of information about the structure all at once.

[0137] The self-driving multi-channel measurement device is waterproof, so that it can operate normally even after being placed outdoors for one year or more.

[0138] An example of placing a sensor or the like outdoors will be described. Fig. 17 is a diagram showing an example of an installation location when placing a sensor outdoors. The structure 100 shown in Fig. 17 is a structure 100 that dams up sand on a slope. In such a structure 100, a strain sensor 305, an impedance sensor 309, and the like are placed as sensors 13. In addition, a wireless communication system 308 is placed as the communication unit 10.

[0139] As shown in FIG. 17, even when each component is placed outdoors, the self-driving multi-channel measurement device operates normally for a long period of time, for example, for two years.

[0140] Long-term monitoring will be described with reference to Figures 18A and 18B. Figure 18A is a diagram showing the results of long-term monitoring. The X axis of the graph shown in Figure 18A represents the number of days. The Y axis of the graph represents temperature, humidity, and resistance value. Figure 18A shows the sensor detection values ​​of the temperature and humidity sensor unit 14, first concrete electrode pair 15, and second concrete electrode pair 16 of Figure 2A. Plot 311 shows the sensor detection value of the first concrete electrode pair 15. Plot 312 shows the sensor detection value of the second concrete electrode pair 16. Plots 313 and 314 show the sensor detection values ​​of the two temperature and humidity sensor units 14.

[0141] The graph shown in Fig. 18A shows the sensor detection values ​​over a three-month period. As shown in Fig. 18A, the sensor detection values ​​are almost the same. This indicates that the sensor 13 continues to operate normally. The sensor 13 can be used for a long period of time.

[0142] The graph shown in Fig. 18B is a diagram showing the correlation between sensor data (sensor detection values) and the relationship with precipitation amount labeling, where the X and Y axes of the graph shown in Fig. 18B both represent resistance values.

[0143] As shown in FIG. 2A , the first concrete electrode pair 15 is placed in the center of the bridge deck 102. The second concrete electrode pair 16 is placed at an end of the bridge deck 102. The X-axis of the graph shown in FIG. 18B represents the sensor detection value of the first concrete electrode pair 15. In other words, the X-axis represents the impedance of the sensor placed in the center of the bridge deck 102. The Y-axis of the graph shown in FIG. 18B represents the sensor detection value of the second concrete electrode pair 16. In other words, the Y-axis represents the impedance of the sensor placed at the end of the bridge deck 102.

[0144] 18B plots the impedance for a sunny day and for days with precipitation of 34 mm / day, 25 mm / day, 22 mm / day, and 5 mm / day. The straight line 320 on the graph in FIG. 18B represents an approximation of the plot for a sunny day.

[0145] The impedance plot on a day with a precipitation rate of 5 mm / day almost overlaps with the approximation line 320. As shown in the circle 322, the impedance plots on days with a precipitation rate of 34 mm / day, 25 mm / day, and 22 mm / day deviate from the approximation line 320. In other words, the relationship between the impedance at the center of the bridge deck 102 and the edge of the bridge deck 102 is different from that on sunny days and days with little precipitation. This suggests that on days with heavy precipitation, water may be dripping from the edge of the bridge. In this way, by sensing the structure 100 at multiple points, abnormalities inside the structure can be detected early.

[0146] (Extraction of correlation from multi-component extraction) In the sensor system 1, specific physical quantities and their correlations can be extracted from multi-component extraction. Extraction of correlation from multi-component extraction will be described with reference to Figs. 19A and 19B. Fig. 19A is a diagram showing daily precipitation amounts. Fig. 19B is a diagram showing impedance after compensation. The X-axis in Figs. 19A and 19B indicates the date. The Y-axis in Fig. 19A indicates precipitation amounts. The Y-axis in Fig. 19B indicates impedance.

[0147] Fig. 19B shows the value obtained by reducing the influence of disturbances and the like from the detected values ​​of multiple sensors using the aforementioned algorithm. That is, Fig. 19B shows the impedance after compensation. Specifically, Fig. 19B shows the ratio of the detected values ​​of two sensors.

[0148] As the values ​​for 1 / 13 (month / day) and 1 / 14 (month / day) in FIGS. 19A and 19C show, the compensated impedance in FIG. 19B correlates with the precipitation amount in FIG. 19A.

[0149] Although the present invention has been described above with reference to the preferred embodiment, it is to be understood that the present invention is not limited to the preferred embodiment described above and that various modifications, variations, and combinations are possible.

[0150] The sensor system of this embodiment is a sensing method that uses a remote monitoring device capable of placing one or more types of sensors at multiple points (several to approximately 100 points) on a large-area structure, and integrates an algorithm to extract and / or compensate for specific physical quantities (temperature, humidity, moisture, strain, cracks, potential, impedance, dielectric constant, electromagnetism, light, ions, pressure, vibration, magnetism, etc.) and correlations between multiple physical quantities from sensor information from multiple points. The sensor system of this embodiment suppresses the "dissipation of physical quantities due to changes over time," a problem that arises during field demonstrations. Furthermore, the sensor system of this embodiment achieves this using an algorithm without the need for additional equipment. In other words, the sensor system of this embodiment can efficiently extract feature quantities from sensor detection values.

[0151] The sensor system of this embodiment realizes long-term outdoor monitoring, something that was not possible with conventional electronics. As a result, it becomes possible to visualize and quantitatively evaluate the condition of structures when needed, allowing for timely timing of repairs or replacements of buildings and other structures. Furthermore, it contributes to reducing the amount of inspections required for excessive safety management and the amount of concrete used for repairs, thereby achieving optimization in terms of both energy consumption and human resources. In other words, the sensor system of this embodiment embodies technology for maintaining and managing structures over decades with less labor and at low cost.

[0152] (1) A sensor system comprising: a plurality of flexible sheet-like sensors arranged on a structure and detecting values ​​indicating the state of the structure; and an analysis unit that, when the values ​​detected by the sensors are taken as sensor detection values, analyzes the sensor detection values; when the analysis unit takes as an analytic value a value generated by analyzing the sensor detection values, the analysis unit performs at least one of the following: setting at least one value among the sensor detection values ​​as a reference value and generating the analytic value based on the reference value; generating the analytic value by performing statistical analysis of the same or multiple types of feature quantities among the sensor detection values; and generating the analytic value by referring to data indicating the relationship between the sensor detection values ​​and the state of the structure; and the sensor system wherein the analytic value contains less external disturbance influence than the sensor detection values.

[0153] (2) The sensor system according to (1), wherein the analysis unit is configured to: when at least one value among the sensor detection values ​​is set to a reference value, generate the analytic value by at least one of determining a difference between the reference value and the sensor detection value other than the reference value, and determining a ratio of the sensor detection value other than the reference value to the reference value; when performing statistical analysis on the same or multiple types of feature quantities among the sensor detection values, generate the analytic value by extracting feature quantities of the analytic values ​​by principal component analysis; when referring to status data indicating a relationship between the sensor detection values ​​and the status of the structure, generate the analytic value by referring to analysis data or analytical data of at least one of an expert's inspection and an image; and output at least one of the feature quantity and soundness as a final output; or process at least one of the feature quantity data and the status data as input using a time-series data analysis technique that generates a predictive model in real time, and output at least one of the predicted feature quantity and soundness.

[0154] (3) The sensor system described in (2), wherein the reference value is generated based on the sensor detection value acquired by the sensor placed either at a position in a healthy state of the structure or at a position in a healthy state simulating the structure.

[0155] (4) The sensor system according to any one of (1) to (3), wherein the sensor exhibits flexibility in at least one of bending and stretching, and when stretchable, has at least a stretchability of 10% strain or a durability of 10% strain or more and 50% strain or less, and when bending, has at least a durability of a bending radius of 1 mm or more and 10 cm or less, and the thickness of the substrate or sealing material is 1 μm or more and 1 mm or less.

[0156] (5) The sensor system according to any one of (1) to (4), wherein the sensor is at least one of a strain sensor, a vibration sensor, a pressure sensor, a tilt sensor, an acceleration sensor, a temperature sensor, a humidity sensor, a water level sensor, a rainfall sensor, an ion sensor, a corrosion sensor, a moisture sensor, a potential sensor, and an impedance sensor.

[0157] (6) A sensor system according to any one of (1) to (5), which is configured by a sheet-shaped or printable electronic circuit that incorporates an analog signal processing unit, a battery unit, an intermittent operation control unit, a central control unit, and a data storage unit, and further incorporates at least one of the analysis unit, communication unit, output unit, light-emitting unit, wireless power supply, and self-sustaining power supply, or the sheet-shaped or printable electronic circuit has flexibility in at least one of bending, stretching, and twisting.

[0158] REFERENCE SIGNS LIST 1 sensor system 2 sensor section 3 processing section 4 central control section 5 sensor driving section 6 battery power section 7 intermittent operation control section 8 power supply control section 9 CPU 10 communication section 11 analog signal processing section 12 data storage section 13 sensor 14 temperature and humidity sensor section 15 first concrete electrode pair 16 second concrete electrode pair 17 third concrete electrode pair 18 concrete rebar connection section 19 concrete electrode 20 first strain sensor 21 second strain sensor 22 thermistor 25 impedance sensor section 26 self-potential sensor section 27 strain sensor section 28 thermistor section 30 sine wave generating section 31 impedance calculation section 32 multiplexer 35 sine wave 36 waveform 37 frame 38 impedance measurement IC 40 relay 41 strain sensor driving section 42 thermistor driving section 45 ADC unit 46 Multiplexer 48 Analysis unit 50 First battery power supply unit 51 Second battery power supply unit 53 Lithium battery 54 Charging control IC 56 Thin (small capacity) lithium battery 57 Low-power charging control IC 58 Step-up DC-DC 59 Solar cell 61 FET (power switch) 62 Flip-flop 63 Voltage determination unit 64 Timer IC 65 Button battery 70 FET (power switch) 71 I / O expansion 75 First communication unit 76 Second communication unit 77 Third communication unit 80 Wi-Fi communication unit 81 Sigfox communication unit 83 LTE (registered trademark)-M SIM 84 SIM conversion IC 85 Communication CPU 100 Structure 102 Bridge deck 104 Frame 106 Concrete surface 108 Rebar 120 3.8V line 122 Start 123 End of sensor measurement 124 End of Sigfox communication 126 One measurement 127 Measurement period 128 Power supply stop period 129 Charging period 130 Cycle 150 Output unit 151 Light-emitting unit 152 Wireless power supply unit 153 Standalone power supply unit 220 Crack 250 AFE circuit 251 Selector 1 to X 252 Amp. 253 ADC 254 CPU 255 Selector 1 to 8 301 Water level / rainfall sensor 302 Tilt sensor 303 Vibration sensor 304 Rebar corrosion sensor 305 Strain sensor 306 Soil moisture sensor 308 Wireless communication system 320 Approximate line.

Claims

1. A sensor system comprising: a plurality of flexible sheet-like sensors that are arranged on a structure and detect values ​​that indicate a state of the structure; and an analysis unit that, when the values ​​detected by the sensors are taken as sensor detection values, analyzes the sensor detection values, wherein, when the analysis unit takes a value generated by analyzing the sensor detection values ​​as an analytic value, the analysis unit performs at least one of the following: setting at least one value among the sensor detection values ​​as a reference value and generating the analytic value based on the reference value; generating the analytic value by performing statistical analysis of the same or multiple types of feature quantities among the sensor detection values; and generating the analytic value by referring to data that indicates the relationship between the sensor detection values ​​and the state of the structure, and wherein the analytic value contains less disturbance influence than the disturbance influence included in the sensor detection values.

2. The sensor system according to claim 1, wherein the analysis unit generates the analytic value by at least one of the following: when at least one value among the sensor detection values ​​is set as a reference value, determining a difference between the reference value and the sensor detection value other than the reference value, and determining a ratio of the sensor detection value other than the reference value to the reference value; when performing statistical analysis of the same or multiple types of feature values ​​among the sensor detection values, generating the analytic value by extracting feature values ​​of the analytic values ​​by principal component analysis; when referring to condition data showing the relationship between the sensor detection values ​​and the condition of the structure, generating the analytic value by referring to analysis data or analytical data of at least one of an expert's inspection and an image; and outputting at least one of the feature value and soundness as a final output; or processing at least one of the feature value data and the condition data using a time series data analysis technique that generates a predictive model in real time using an input of the feature value data and / or the condition data, and outputting at least one of the predicted feature value and soundness.

3. The sensor system according to claim 2, wherein the reference value is generated based on the sensor detection value acquired by the sensor placed either at a position in a healthy state of the structure or at a position in a healthy state simulating the structure.

4. The sensor system according to any one of claims 1 to 3, wherein the sensor exhibits flexibility in at least one of bending and stretching, and when stretchable, has at least a stretchability of 10% strain or less, or a durability of more than 50% strain, and when bending, has at least a durability of a bending radius of 1 mm or more and 10 cm or less, and the thickness of the substrate or sealing material includes 1 μm or more and 1 mm or less.

5. The sensor system according to any one of claims 1 to 4, wherein the sensor is at least one of a strain sensor, a vibration sensor, a pressure sensor, a tilt sensor, an acceleration sensor, a temperature sensor, a humidity sensor, a water level sensor, a rainfall sensor, an ion sensor, a corrosion sensor, a moisture sensor, an electric potential sensor, and an impedance sensor.

6. A sensor system according to any one of claims 1 to 5, which is comprised of a sheet-shaped or printable electronic circuit incorporating an analog signal processing unit, a battery unit, an intermittent operation control unit, a central control unit and a data storage unit, and further incorporating at least one of the analysis unit, communication unit, output unit, light-emitting unit, wireless power supply and independent power supply, or the sheet-shaped or printable electronic circuit has flexibility in at least one of bending, stretching and twisting.

Images