JP5075762B2 - Sensor node and sensor network system - Google Patents

Description

  The present invention relates to a sensor network system in which a large number of sensor nodes are connected by a wireless network, data measured by the sensor nodes is collected by a computer, and data analysis is performed. The present invention relates to a sensor network system that prevents communication congestion between sensor nodes by reducing the communication amount of the sensor nodes in a congested communication environment.

  In recent years, development of sensor network systems composed of small wireless sensor nodes (hereinafter referred to as sensor nodes), repeaters (router nodes), base stations, and sensor network servers (hereinafter referred to as management servers) equipped with sensor functions has been promoted. ing. The sensor node acquires data (hereinafter referred to as sensor data) related to the state of a person or a place by a sensor, relays the acquired sensor data in a multi-hop by a relay, and transmits the data to a management server via a base station. The management server executes various processes based on the received sensor data.

  A key device in a sensor network system is a sensor node characterized by small size and low power. Because it is small in size, it can be attached to everything including the environment and people, and because of its low power, it can be operated for several years with a battery without external power supply. Furthermore, since communication is performed wirelessly, the degree of freedom of the installation location of the sensor node is high, and the sensor node can be arranged in a wide range.

  As a characteristic operation in the sensor node, there is an intermittent operation. This drives the necessary hardware only when executing tasks such as sensing and data transmission. When there are no tasks to be executed, peripheral hardware such as sensors and RF are completely stopped, and the microcomputer is also low power. It is an operation to sleep in the mode. By performing the intermittent operation, the sensor node can operate for a long time under a limited battery.

  When the microcomputer expires a predetermined sleep time, the timer unit generates an interrupt, so that the microcomputer returns to the normal operation mode. Then, according to a predetermined procedure, sensing is performed, the sensed data is transmitted, and if there is data addressed to itself, it is received by Polling and the data is processed. When all the tasks to be executed at that time are completed, the computer enters sleep for a predetermined time again. Since the length of the task processing period between the sleep periods is about 1 second at the longest from several tens of milliseconds, the sensor node is almost in the sleep state for that time.

  Since the basic feature of intermittent operation is “sleep when there is no need to execute a task”, various methods can be considered for the method of performing intermittent operation depending on the task scheduling method. For example, when a sensor node is equipped with a plurality of sensors, different sensing cycles may be provided for each sensor. In addition, since sensing, data transmission, and data reception are inherently independent tasks, each task may have an independent operation cycle.

  Use of wireless communication between the sensor node and the management server is necessary from the viewpoint of the degree of freedom of installation location, but inconvenience also occurs. In the case of IEEE802.15.4, which is a low-power wireless communication method used in sensor networks, the communication rate is a maximum of 250 kbps (31.25 kB / s). However, when continuously observing data with a relatively high sampling rate such as vibration and voice, high-rate communication is required. Therefore, as the number of sensor nodes belonging to one wireless network increases, the band that can be used by one sensor node is limited.

  In order to effectively use the bandwidth of the wireless network, a technology is known that reduces the amount of communication by thinning out and transmitting data with small changes among the data measured by the sensor node and transmitting only the data difference. (For example, refer to Patent Document 1).

  In addition, when the sensor node calculates and transmits a histogram using a plurality of observation values and the server determines that the received histogram is an abnormal value, the plurality of observation values used by the sensor node to calculate the histogram Is known (for example, see Patent Document 2).

JP2003-122796 JP2002-256469

  The sensor network system as an object of the present invention is to install a large number of sensor nodes on a measurement object, acquire sensor data at a high frequency, and perform analysis. For example, it is an abnormality diagnosis / preventive maintenance system by monitoring vibrations of equipment such as pipes, pumps, and motors. Specifically, battery-driven vibration sensor nodes that measure vibrations of about 400 Hz to 1 kHz are always installed at 100 to 10,000 locations such as pipes and equipment to detect anomalies.

  When performing preventive maintenance of equipment using such vibration observation values, Fourier transform is used. Since Fourier transform is a technique for obtaining a frequency spectrum from continuous observation values, continuous observation values for a certain period are required. Furthermore, the management server needs to collect the continuous observation values as actual data that has not been subjected to thinning processing or the like. Therefore, for example, data thinning cannot be performed as disclosed in Patent Document 1. Further, even if data thinning is performed, in a sensor network system including a large number of sensor nodes having a large data transmission amount per unit time, there arises a problem that the communication band becomes tight as the number of sensor nodes increases.

  Further, as disclosed in Patent Document 2, a sensor node transmits a statistical value calculated from an observation value during a certain period to the server, requests an observation value during a period when the server determines that the abnormality is detected, and the sensor node performs the observation. When transmitting a value, the amount of communication increases due to the transmission of statistical values and the transmission of observation value requests. Furthermore, since the sensor node cannot sleep until receiving a request, there arises a problem that power consumption increases.

  In order to solve the above-described problems, an object of the present invention is to provide a sensor network system that can secure a communication band of a wireless network while including a large number of sensor nodes that measure data with a high sampling rate.

  Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows. A sensor that acquires observation values at predetermined intervals, an approximate calculation unit that calculates a feature value from the observation values acquired during a predetermined period, a transmission determination unit that determines whether the feature value exceeds a predetermined threshold, and a feature A wireless communication unit that transmits an observation value acquired during the predetermined period to a server when the amount exceeds the predetermined threshold.

  Further, in the sensor network system having a sensor node and a server, the sensor node includes a sensor that acquires observation values at predetermined intervals, an approximate calculation unit that calculates a feature value from the observation values acquired in a predetermined period, A sensor having a transmission determination unit that determines whether or not a feature amount exceeds a predetermined threshold, and a wireless communication unit that transmits an observation value acquired in a predetermined period to the server when the feature amount exceeds a predetermined threshold It is a network system.

  ADVANTAGE OF THE INVENTION According to this invention, the communication band of a wireless network can be ensured, enabling the monitoring of the monitoring object by a sensor node.

  In particular, even when a large number of sensor nodes that perform observation at a predetermined frequency are provided, the communication network amount is reduced by reducing the amount of communication data, and a sensor network system having a large number of sensor nodes. Can be operated stably.

  Before describing embodiments of the present invention, features of the present invention will be described. The feature of the present invention is that observation is performed only when a sensor node calculates a feature value from a plurality of observation values acquired within a predetermined period, determines whether transmission is necessary using the feature value, and determines that transmission is required. Send value. As a result, the transmission amount of the sensor node can be reduced to secure the communication band of the wireless network, and the server can collect data (observation values) necessary for analysis. Here, the feature quantity is a quantity expressed by quantifying the feature of the observed value, and is a quantity that can be used as a criterion. A plurality of observation values are used to calculate a single data or a plurality of data having a data amount smaller than a plurality of observation values and use them as feature amounts.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram of a configuration of a sensor network system according to an embodiment of the present invention.

  The sensor network system includes a management server AP1, a gateway GW1, a router node RT1, a sensor node SS1, a management computer 101, a client computer 102, a wired sensor 103, an RFID reader 104, and a LAN (Local Area Network) NW2.

  The LAN NW2 connects the management server AP1, the gateway GW1, the management computer 101, the client computer 102, and the wired sensor 103 to each other.

  The sensor node SS1 is attached to a monitoring target, measures measurement data such as vibration and sound at a predetermined sampling period, and transmits the measurement data to the base station via the wireless network NW1 adopting IEEE802.15.4 as a physical layer. Distributed and installed in the sensor network system. Although nine sensor nodes are installed in the sensor network system of this explanatory diagram, any number may be installed. The sensor node includes a ZigBee communication unit 122, a task manager 124, a sensor controller 125, and a power manager 126. The ZigBee communication unit 122 communicates with the gateway GW1 and the router node RT1 using the ZigBee protocol. The task manager 124 processes tasks such as activation and sleep for intermittent operation, and instructions from the user. The sensor controller 125 acquires an observation value from the sensor. The power manager 126 performs power control during activation and sleep. Details regarding the sensor node will be described with reference to FIG.

  The router node RT1 includes a ZigBee communication unit 122 and a routing manager 121. The ZigBee communication unit 122 communicates with the gateway GW1, the sensor node SS1, or another router node using the ZigBee protocol. The routing manager 121 determines a transfer destination of information received from the outside. In the sensor network system of this explanatory diagram, seven router nodes are installed, but any number may be installed.

  The router node RT1 receives the environmental information observed by the sensor node SS1 or the request issued from the management server AP1 and transfers it to the gateway GW1, the sensor node SS1, or another router node RT1. The gateway GW1 includes a LAN communication unit 120, a routing manager 121, a ZigBee communication unit 122, and a PAN control unit 123.

  The LAN communication unit 120 communicates with the management server AP1 via the LAN NW2. The routing manager 121 determines a transfer destination of information received from the outside. The ZigBee communication unit 122 communicates with the router node RT1 or the sensor node SS1 using the ZigBee protocol. The PAN control unit 123 controls a PAN (a wireless communication network including a sensor node, a router node, and a gateway) configured from the gateway GW1.

  The gateway GW1 receives the environmental information observed by the sensor node SS1 or a request issued from the management server AP1, and transfers it to the management server AP1, the router node RT1, or the sensor node SS1.

  The management server AP1 communicates with the base station GW1, the management computer 101, the client computer 102, and the wired sensor 103 via the LAN communication unit 120 and the LAN NW2. In addition, it receives observation values from the sensor node via the base station and controls the sensor node. The management server includes an analysis function 110, an alarm function 111, and a management function 112. The analysis function 110 analyzes the observation value obtained from the sensor. The alarm function 111 notifies the user if there is any abnormality as a result of analyzing the observed value. The management function 112 performs network configuration management and sensor node status management. Further, the management server includes a database HistoryDatabase 116 that stores observation values and analysis results.

  The management computer 101 is operated by an administrator of the sensor network system. The management computer 101 sends various requests to the management server AP1 in response to the administrator's operation.

  The client computer 102 is operated by a client of the sensor network system. The client computer 102 executes various applications. Further, the client computer 102 receives environment information observed by the sensor node SS1 from the management server AP1. The client computer 102 performs various processes based on the received environment information.

  The wired sensor 103 observes environmental information. Then, the wired sensor 103 transmits the observed environment information to the management server AP1 via the LAN NW2. The wired sensor is a group of sensors connected to the network NW2 by wire, for example, a thermocouple temperature sensor. The RFID 104 reader reads the ID described in the RFID Tag and notifies the server through the network NW2.

  The sensor node SS1 includes an acceleration sensor that measures three-axis acceleration in the X-axis, Y-axis, and Z-axis directions as a vibration sensor. Moreover, you may provide the temperature / humidity sensor which measures temperature and humidity, the microphone which measures a sound, and the illumination intensity sensor which measures a brightness. An example of the hardware configuration of the sensor node is shown in FIG.

  The sensor node SS1 includes a sensor 201, an A / D converter 202, a microcomputer 203, a wireless communication unit 204, a display unit 205, a nonvolatile memory 206, a read-only memory 207, a real-time clock 208, a battery 209, and a button 212. The wireless communication unit 204 corresponds to the ZigBee Communicator 122 in FIG.

  The output of the acceleration sensor for each axis is converted from an analog signal to a digital signal by the A / D converter 202. The observation value converted into the digital signal is read into the microcomputer 203 that performs arithmetic processing, and is written into the nonvolatile memory 206 such as a flash memory together with the time information of the Real-Time Clock 208.

  The microcomputer 203 performs various processes described in FIG. 3 by reading the program stored in the nonvolatile memory 206 into the memory 210 and executing it at the time of activation. When the observation value read from the A / D converter 202 is stored in the memory 210 and observed a predetermined number of times, a transmission determination described later is performed from the observation value. Then, it is stored in a packet as necessary, and is wirelessly transmitted from the wireless communication unit 204 to the relay station or the base station via the antenna Antenna. These series of tasks are managed by a task manager 124 operating on the microcomputer 203.

  The display unit 205 is an output device for displaying information to the user, and can display the latest observation value measured by the sensor 201. At this time, in order to reduce power consumption, nothing is displayed during normal operation, and the latest measurement value is displayed only when a specific operation (for example, button 212 is pressed) is performed.

  The sensor node uses, in hardware, two operations, a sleep state in which the power other than the Real-Time Clock 208 is turned off and a start-up state in which the power of all the circuits is turned on at regular intervals. When the sensor node enters a sleep state, the sensor node can operate while maintaining time information with a low power of 5 μA.

  An example of functional elements of the sensor node SS1 is shown in FIG. The sensor node according to the present embodiment is characterized in that an approximate calculation is performed using observation values acquired within a predetermined period, and it is determined whether or not an observation value transmission is necessary. As a result, the data transmission amount of the sensor node can be reduced. Details of the approximate calculation will be described later.

  The sensor data acquisition unit 301 converts the sensors SnX to SnZ of each axis into digital signals. The sensor data acquisition unit 301 corresponds to the SensorController 125 in FIG. The nonvolatile memory storage unit 305 adds a time stamp indicating the observation time to the observation value converted into the digital signal, and stores it in a nonvolatile memory such as a flash memory. The approximate calculation unit 302 performs an approximate calculation of the observation value in the period from the observation value converted into a digital signal every time a predetermined period (for example, 1 second) is reached. The transmission determination unit 303 determines whether or not the observation value transmission is necessary from the approximate value. The wireless communication unit 304 performs transmission / reception with the base station via a wireless network. The request reception unit 306 receives a request (command) from the management server. The data discarding unit 307 deletes the observation value when the received request is a request to discard the data in the nonvolatile memory. When the request is a transmission determination threshold setting request, the threshold setting unit 309 notifies the transmission determination unit 303 of the threshold included in the threshold setting request and stores the threshold in the nonvolatile memory. When the received request is a request for acquiring an observation value in the memory, the observation value request processing unit 309 acquires the approximate observation value from the memory and transmits it to the data generation unit 308. In the data generation unit 308, the observation value is stored in a packet and transmitted to the base station via the wireless communication unit 304. An example of the hardware configuration of the management server AP1 is shown in FIG. The management server AP1 includes a CPU 401 that performs calculations, an external communication unit 402, a power supply 403, a hard disk drive 404, a keyboard 405, a display 406, and a memory 407. The external communication unit 402 corresponds to the LAN Communicator 120 in FIG.

  The management server AP1 communicates with the wired network NW2 via the external communication unit 402, transmits commands to the base station GW1, and the observation values from the sensor node SN1 collected by the base station GW1 via the relay RT1. Receive. The management server has a management function 112 for managing the wireless network configuration of base stations, repeaters, sensor nodes, intermittent operation settings of sensor nodes, and the like. The CPU 401 reads a program stored in the memory 407 and executes each process described with reference to FIG. For example, data such as observation values acquired in accordance with program instructions is processed, data is stored in the hard disk drive 404, or displayed on the display 406. Further, the CPU 401 reads the history stored in the History Database 116 on the hard disk drive 404 and displays it on the display 406 in response to a user request input from the keyboard 405. In addition, the CPU 401 interprets the user command input from the keyboard 405 and distributes it to the base station GW1 through the external communication unit 402.

  Examples of functions of the management server AP1 include an event reception unit 502, an event delivery unit 503, an observation value storage unit 504, a display unit 505, a detailed analysis unit 506, a user request reception unit 507, and a command creation unit 508 as shown in FIG. And a command issuing unit 509. The detailed analysis unit 506 corresponds to the analysis function 110 in FIG.

  The management server AP1 receives the observation value 501 from the sensor node SS1 at the event reception unit 502 of the external communication unit 402. The received observation value 501 is delivered to and stored in the observation value storage unit 504 on the hard disk drive 404 through the event delivery unit 503, or delivered to the display unit 505 on the display 406. In addition, when the observation value 501 is analyzed in detail, it is analyzed after being delivered from the event delivery unit 503 to the detail analysis unit 506. The analysis result is displayed on the display unit 505, and an alarm is issued to the user through the alarm function 111 as necessary.

A request from the user to the sensor node is input to the user request receiving unit 507 via the keyboard 406. The user request is interpreted as a command for the sensor node by the command creation unit 508, and issued from the command issuing unit 509 of the external communication unit 402 to the sensor node.
FIG. 6 shows an example of observing piping vibration as an example of an application using the above sensor network system.

  The pipe 603 is fixed to the building by the support 607, but vibrates due to the influence of the connected pump 601 and the influence of the fluid flowing therethrough. In the case of a steel material, a stress such as a crack is caused by repeatedly applying stress exceeding a certain value (fatigue limit) among stresses generated by this vibration. Therefore, it is necessary to measure the vibration and evaluate the stress. As shown in FIG. 6, the vicinity of the valve 602, the branching portion 604, the pipe joint 606, and the bent portion 605 generate a drift, and thus the vibration is particularly large. Therefore, a plurality of sensor nodes SS1 are installed, vibrations are measured, and stress is evaluated.

  The sensor node of the present embodiment is characterized in that stress is evaluated by approximating a general physical model with statistical values after measuring vibration. Thereby, stress can be evaluated with a small amount of calculation, and stress can be evaluated at high speed even for a sensor node having low performance such as a microcomputer. Furthermore, evaluation according to a physical model is possible, and determination with high accuracy can be performed.

Generally, piping stress analysis uses a method for evaluating vibration characteristics based on acceleration power spectral density PSD (Power Spectrum Density). A formula for calculating the vibration stress σ [N / mm ² ] of the pipe, which is a physical model, is shown in Formula (1).

  In addition, a formula for calculating the power spectral density for each analysis frequency is shown in Formula (2).

As disclosed in Patent Document 2, PSD is calculated by Fourier transforming an observed value. The amount of calculation required to perform the Fourier transform is on the order of NlogN, where N is the number of observation data points. That is, in order to Fourier-transform 512 points, after 512 observations, a calculation amount of 512 × log512 = 1388 order is required.
On the other hand, in this embodiment, when performing stress analysis of a structure, evaluation is performed using a power spectrum at the primary natural frequency as a simple evaluation. In this case, since f, E, e, L, and Δf can be constants, the stress can be approximated by Equation (3) and simply evaluated.

In Equation (3), Σx _i ² can be calculated sequentially each time observation is performed. Accordingly, the amount of calculation necessary for obtaining σ is about one order because division only needs to be performed once. As described above, the evaluation formula (3) used in the present embodiment has a smaller calculation amount than the physical model (1) and can be evaluated according to the physical model.

  FIG. 9 shows an example of a flowchart of the observation value transmission determination. Here, the sensor node calculates the above approximate expression, and transmits an observed value when the approximate value exceeds the threshold value, and does not transmit when the approximate value does not exceed the threshold value. As a result, the calculation amount of the sensor node can be reduced and the communication amount can be reduced. Furthermore, the server can collect actual data of observations that require detailed analysis.

  When the sensor node SS1 is activated, the observation is performed (901), the observation value is stored in the nonvolatile memory (902), and the square of the observation value is calculated and added (903). It is determined whether the number of observations exceeds a specified number (for example, 512) (904). If not, observation 901 is performed again, and the square of the observation value is added (903). The specified number of times is a value determined by the details of detailed analysis performed by the management server, and is specified by the management server. In addition, by performing sequential calculation that adds squares after observation, it is possible to reduce the amount of calculation at one time and store it in the memory 210 as compared with calculations that perform processing after a specified number of times. The amount of data can be reduced to one observation and two data for storing the calculation results. As a result of the determination, if the number of observations exceeds the specified number, an approximate value is calculated by equation (3) (905). It is determined whether or not the approximate value exceeds the set threshold value (906). If the threshold value is exceeded, the observed value is transmitted to the server (907) and the process ends (908). If the approximate value is less than or equal to the set threshold, the observation value is not transmitted and the process ends (908).

  The threshold value may be set in advance by the user, or can be determined and set by the sensor node as described later in the second embodiment. Also, the server can set a threshold according to the transmission state of the observation value from the sensor node. For example, when the observed value exceeding the threshold is not observed for a predetermined time in the server, the threshold can be set low within a range where the communication band is not tight.

  Further, when a plurality of sensor nodes are installed at different locations, the threshold value can be individually determined for each sensor node. That is, the threshold value is set according to the installation location of the sensor node, and the sensor node transmits the observation value only when the approximate value σ exceeds the threshold value. As a result, even if the vibration level at the installation location is different, the transmission amount of the observation value transmitted by each sensor node is the same, or the transmission amount of the sensor node installed at the location where you want to observe intensively, such as near the pump It becomes possible to set many. In this case, the user may set the threshold in advance according to the installation location. Alternatively, it is also possible to set the server node by knowing the installation location of the sensor node and transmitting a threshold setting command to each sensor node installed at the installation location.

  As described above, since the observation value continuously observed at the sensor node SS1 is determined to be transmitted every predetermined number of observations and is not transmitted unless the threshold is exceeded, it is continuously transmitted. Even if there are a large number of sensor nodes SS1 to SSn to be observed, it is possible to prevent the communication band of the wireless network NW1 from becoming tight and to secure a stable communication environment.

  Here, since the approximate value is not required in the server analysis, the sensor node does not need to transmit the approximate value from the viewpoint of reducing the communication amount and the power consumption. However, as shown in FIG. 11, when the sensor node calculates 905 an approximate value and determines that the observation value is not transmitted by the transmission determination 906, the calculated approximate value can be transmitted. As a result, even if measurement of an observed value that does not exceed the threshold value continues, an approximate value is transmitted from the sensor node, so the server determines whether the sensor node is alive (determining whether the sensor node is functioning normally) Can be performed.

  Next, FIG. 7 shows the relationship between the vibration data observed at the sensor node SS1 and the vibration data acquired at the management server AP1. In FIG. 7, 701 indicates vibration data observed at the sensor node SS1, 702 indicates an approximate value of stress calculated at the sensor node, and 703 indicates the state of vibration data acquired by the management server AP1. That is, all the observed values are recorded in the sensor node, and only the observed value whose approximate value exceeds the threshold is recorded in the management server.

  The sensor node SS1 continuously performs observation at a specified sampling period (for example, 200 Hz) 711, and calculates an approximate value 721 by using an observed value for each specified number of times 712 (for example, 512 times). When the calculated approximate value 721 exceeds the determination threshold value 722, the observed value 711 of the section 712 that has exceeded is transmitted to the management server AP1. Therefore, the management server AP1 acquires only vibration data that exceeds the threshold 722 as indicated by 703.

  Since the sensor node SS1 stores the observed value 711 in the non-volatile memory 206, the management server AP1 should inquire and acquire the sensor node SS1 when vibration data for the section (713) that does not exceed the threshold 722 is required. Is possible.

  Observation of the sensor nodes SS1 to SSn and communication with the base station are performed as shown in the sequence diagram of FIG.

  The sensor node SS1 observes a predetermined number of times (713), and stores the observed value 801 in the nonvolatile memory (902). When the predetermined number of times is exceeded, an approximate value is calculated (905). It is determined whether the calculated approximate value exceeds the threshold value (906). If not, the observation value is not transmitted and the observation is performed again. When the approximate value calculated from the predetermined number of observation values (712) exceeds the threshold, the observation value is transmitted (802). If an observation value that has not been transmitted is necessary, the management server AP1 transmits an observation value request to the sensor node SS1 via the base station (803). The sensor node SS1 acquires the observation value stored in the nonvolatile memory (804), and transmits the observation value to the base station GW1 (801).

  Next, as a second embodiment, an example in which the sensor node determines and resets the threshold will be described with reference to FIG. In this embodiment, in order to prevent communication congestion, the sensor node monitors the transmission amount of the observation value, and adjusts the transmission amount by changing the threshold when the transmission amount increases or decreases. It is characterized by that.

  FIG. 10 shows a flowchart in which the sensor node determines and changes the threshold value. The flow for performing the approximate value calculation 1002 is omitted because it is the same flow as the observation 901, the storage 902 in the non-volatile memory, the sequential calculation 903, the overrun determination 904, and the approximate value calculation 905 in FIG.

  When the sensor node is activated (1001), the approximate value is calculated in the same flow as the approximate value calculation of FIG. 9 (1002), and the number of calculations is added (1003). It is determined whether or not the calculated approximate value exceeds the threshold value (1004). If the calculated approximate value exceeds the threshold value, the observation value is transmitted to the server and the number of transmissions is added (1005). Thereafter, it is determined whether or not the number of approximate value calculations exceeds a predetermined number (for example, 10 times) (1006). If not, the process ends (1013) and approximate value calculation is performed again (1002). ). If the predetermined number is exceeded, the transmission rate is calculated (1007). Based on the calculated transmission rate, it is determined whether or not to recalculate the threshold (1008). FIG. 10 shows an example in which recalculation is performed on condition that the transmission rate is greater than 0.6 or the transmission rate is 0.

  The transmission rate is the ratio of the number of transmissions to the approximate number of calculations (transmission determination number). When the transmission rate is larger than the determination value (for example, 0.6), it is determined that the transmission is excessive, and when it is smaller than the determination value (for example, 0), it is determined that the transmission is insufficient. The judgment value of the transmission rate can be set for each sensor node. For example, for sensors installed in places that you want to monitor intensively, such as near the pump, if you set it to be recalculated if it is greater than 0.9, observation It becomes possible to collect more values.

  When it is determined that there is excessive transmission or insufficient transmission, an average value of approximate values for a predetermined number of times is calculated (1009) and set as a new threshold (1010). Thereafter, the number of calculations is reset to 0 (1011), the number of transmissions is reset to 0 (1012), and the process ends (1013).

  In this way, by resetting the threshold value of the approximate value based on the determination result of the transmission rate, the transmission rate of each sensor node can be kept constant, and even if one node has excessive transmission, the communication band is tight. Stable observation value collection is possible.

  The transmission rate calculation and threshold recalculation described above can also be performed by a server. The sensor node determines whether or not the approximate value exceeds the threshold (1004). If not, the sensor node transmits only the approximate value to the server to notify “not over”. When the server receives the approximate value a predetermined number of times (for example, 10 times), the server calculates the transmission rate (the ratio of the number of times the observed value is received to the number of times the feature value is received), and the transmission rate exceeds the judgment value (for example, 0.6) In this case, the average of the approximate values for 10 times is calculated. Then, a command for resetting the average value as a threshold value is created and transmitted to the sensor node through the command issuing unit 509. The sensor node sets an approximate value threshold based on the received command.

  Thereby, although it is disadvantageous to transmit the approximate value to the server from the viewpoint of the communication amount and the power consumption of the sensor node, the power consumption due to the calculation of the transmission rate and the recalculation of the threshold can be reduced.

  Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications can be made, and it is possible to appropriately combine the above-described embodiments. It will be understood by the contractor.

1 is an example of a diagram illustrating a configuration of a sensor network system to which the present invention is applied in Embodiment 1. FIG. 1 is an example of a block diagram illustrating a hardware configuration of a sensor node in Embodiment 1. FIG. FIG. 2 is an example of a block diagram illustrating a functional configuration of a sensor node in the first embodiment. 1 is an example of a block diagram illustrating a hardware configuration of a management server in Embodiment 1. FIG. 1 is an example of a block diagram illustrating a functional configuration of a management server in Embodiment 1. FIG. An example of the figure which shows the piping vibration monitoring system which is an application which applies this invention in Example 1. FIG. An example of the figure which shows the relationship between the measurement data in Example 1, an approximate value, and the data which a server acquires. An example of the sequence diagram which shows the communication procedure of the sensor node and base station in Example 1. FIG. 6 is an example of a flowchart of transmission determination performed by the sensor node in the first embodiment. An example of the flowchart in which the sensor node in Example 2 resets a threshold value. FIG. 6 is another example of a sequence diagram illustrating a communication procedure between the sensor node and the base station in the first embodiment.

Explanation of symbols

101: management computer, 102: client computer, 103: wired sensor, 104: RFID reader, 201: sensor, 204: wireless communication unit, 205: display unit, 301: sensor data acquisition unit, 302: approximate calculation unit, 303: Transmission determination unit, 304: wireless communication unit, 305: nonvolatile memory storage unit, 306: request reception unit, 307: data discard unit, 308: data generation unit, 309: threshold setting unit

Claims (12)

A sensor for obtaining observation values at predetermined intervals;
An approximate calculation unit for calculating a feature amount from the observed value acquired in a predetermined period;
A transmission determination unit that determines whether or not the feature amount exceeds a predetermined threshold;
A sensor node comprising: a wireless communication unit that transmits an observation value acquired during the predetermined period to a server when the feature amount exceeds the predetermined threshold. The sensor node according to claim 1,
The approximate calculation unit is a sensor node that calculates the feature amount using a power spectrum at the primary natural frequency of the observed value. The sensor node according to claim 1,
A sensor node in which, when the feature amount does not exceed the predetermined threshold, the wireless communication unit does not transmit an observation value acquired during the predetermined period to the server. The sensor node according to claim 1,
When the feature amount does not exceed the predetermined threshold, the wireless communication unit transmits the feature amount to the server. The sensor node according to claim 1,
Holds the number of calculation times of the feature amount and the number of times of transmission of observation values acquired during the predetermined period,
A sensor node having a threshold resetting unit that resets a threshold using the average value of the feature amounts when a ratio of the number of transmissions to the number of calculations exceeds a first determination value. The sensor node according to claim 5, wherein
When the ratio of the number of transmissions to the number of calculations does not exceed a second determination value that is smaller than the first determination value, the threshold resetting unit resets the threshold using the average value of the feature amounts . The sensor node according to claim 1,
When multiple sensor nodes are installed in the equipment to be observed,
A sensor node in which the threshold value is set for each of the sensor nodes according to the installation location of the sensor node in the facility. A sensor network system having a sensor node and a server,
The sensor node is
A sensor for obtaining observation values at predetermined intervals;
An approximate calculation unit for calculating a feature amount from the observed value acquired in a predetermined period;
A transmission determination unit that determines whether or not the feature amount exceeds a predetermined threshold;
A sensor network system comprising: a wireless communication unit configured to transmit an observation value acquired during the predetermined period to the server when the feature amount exceeds the predetermined threshold. The sensor network system according to claim 8,
The approximate calculation unit is a sensor network system that calculates the feature amount by using a power spectrum at the primary natural frequency of the observed value. The sensor network system according to claim 8,
The sensor network system in which the wireless communication unit transmits the feature value to the server when the feature value does not exceed the predetermined threshold. The sensor network system according to claim 10, wherein
The server
An external communication unit that receives the observed value and the feature value acquired during the predetermined period;
An analysis unit that calculates a threshold value using an average value of the feature amount when a ratio of the number of times of receiving the observation value to the number of times of receiving the feature amount exceeds a first determination value;
The external communication unit transmits a threshold value calculated by the analysis unit to the sensor node. The sensor network system according to claim 11, wherein
If the ratio of the number of times the observed value is received to the number of times the feature value is received does not exceed the second determination value that is smaller than the first determination value, the analysis unit uses the average value of the feature value to set a threshold value Sensor network system to calculate

Images