JP2007193511A - Sensor terminal - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To preferentially transmit important information to a server without needing complicated setting of parameter or the like. <P>SOLUTION: In the present invention, a sensor terminal 2 determines the priority order of sensor information transmitted from a sensor, and transmits it to an accumulation server 1 which performs detection of event. When determination of priority order is performed, the determination is performed by a learning function using past sensor information and event occurrence information without setting of a parameter. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a sensor information collection system including a sensor terminal and a server that determines an event condition based on sensor information from the sensor terminal, and more particularly to a sensor terminal used in the sensor information collection system.

  2. Description of the Related Art Conventionally, a sensor information collection system has been developed in which sensor information from a sensor terminal is collected by an accumulation server and an event condition is determined based on the collected sensor information. In such a sensor information collection system, when the number of sensor terminals and the amount of sensor information to be transmitted increase, the load on the storage server and the network increases, and it becomes impossible to quickly determine the occurrence of an event.

  In order to solve such a problem, a method has been devised in which important sensor information is determined and preferentially transmitted by setting parameters in advance in the sensor terminal (see Patent Document 1).

Japanese Patent Application Laid-Open No. 2003-333854

However, in the method of setting parameters in advance in the sensor terminal, it is difficult to calculate parameters when the number of sensor terminals is large or the conditions of event occurrence are complicated, and complicated parameter settings are required. There was a problem of becoming.
The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a sensor terminal that can transmit important information to a server preferentially without the need for complicated parameter settings. To do.

  According to the first aspect of the present invention, in a sensor terminal that transmits sensor information to a server that performs event processing based on sensor information, receiving means that receives sensor information including sensor output values from a plurality of sensors; Event occurrence probability calculating means for calculating an event occurrence probability in the server from the received plurality of sensor information, past sensor information and past event occurrence information, and the calculated event occurrence probability is a predetermined occurrence probability And a priority control means for transmitting the received plurality of sensor information to the server in preference to the other sensor information.

According to a second aspect of the present invention, in the first aspect, the event occurrence probability calculating means comprises:
Compares past sensor information and the received multiple sensor information with the sensor output value at the most recent sampling time, and samples the status information indicated by increase, no change, or decrease, and past event occurrence information A state transition table stored in association with each time and sensor, and a pattern of the state information generated for each of a plurality of windows set in the state transition table, and an event occurrence probability corresponding to the state information pattern A plurality of probability tables stored therein, and means for determining event occurrence probabilities in the server for the received plurality of sensor information based on the plurality of probability tables.

  According to the present invention, it is possible to provide a sensor terminal capable of preferentially transmitting important information to a server without performing complicated parameter calculation or the like.

Hereinafter, a sensor information collection system according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is an overall view of a sensor information collection system according to an embodiment of the present invention.
As shown in the figure, the storage server 1 is connected to a plurality of sensor terminals 2 (2-1 to 2-n) via a network. A plurality of sensors 3 (3-1 to 3-n) are connected to each sensor terminal 2 (2-1 to 2-n).

  Each sensor terminal 2 (2-1 to 2-n) includes an event prediction unit 21 (21-1 to 21-n) and a sensor information transmission unit 22 (22-1 to 22-n). The event prediction unit 21 receives sensor information from the corresponding plurality of sensors 3, and performs event prediction processing according to the embodiment of the present invention based on the received plurality of sensor information. This event prediction process will be described later in detail. The sensor information transmission unit 22 transmits the sensor information received from the sensor 3 to the storage server 1.

The sensor information includes information such as sensor identification information (ID), sensor terminal ID, sensing data, sensing data sampling time (or “sampling number”), and the like.
The storage server 1 includes an event detection unit 11, an information storage unit 12, and an event occurrence information feedback unit 13. The information storage unit 12 stores the sensor information transmitted from the sensor information transmission unit 22 of the sensor terminal 2.

The event detection unit 11 performs an event detection process based on the sensor information stored in the information storage unit 12. The information storage unit 12 stores sensor information transmitted from each sensor terminal 2.
When an event is detected by the event detection unit 11, the event occurrence information feedback unit 13 is included in the sensor terminal 2 that has transmitted the sensor information in which the event has occurred and in the sensor information in which the event has occurred. Event occurrence information including sampling time (sampling number) is transmitted. The sensor terminal that has transmitted the sensor information in which the event has occurred can be recognized by the event occurrence information feedback unit 13 because the sensor information includes the identification information of the sensor terminal.

  Here, “event occurrence” means that the storage server 1 takes some action due to the sensor information collected from the sensor 3. The event generation conditions are input to the storage server 1, and the storage server 1 determines whether to generate an event based on the input information.

FIG. 2 is a more specific functional block diagram of the sensor terminal.
As shown in the figure, the sensor terminal 2 includes a sensor information reception unit 31, a sensor information temporary storage unit 32, a probability table calculation / holding unit 33, an event occurrence probability prediction unit 34, an event occurrence information reception unit 35, and a priority control unit. 36 and a sensor information transmitter 37.

The sensor information receiving unit 31 receives sensor information from each sensor 3.
The sensor information temporary storage unit 32 temporarily stores the sensor information received by the sensor information reception unit 31.
The probability table calculation / holding unit 33 calculates and holds a probability table based on the sensor information temporarily stored in the sensor information temporary storage unit 32 and the event occurrence information received by the event occurrence information receiving unit 35. To do. Details of the calculation of the probability table will be described later.

The event occurrence probability prediction unit 34 obtains the event occurrence probability of the sensor information received by the sensor information receiving unit 31 based on the probability table held in the probability table calculation / holding unit 33.
Here, the “event occurrence probability” means the probability that the storage server 1 will generate an event when the sensor terminal 2 sends sensor information to the storage server 1. Since there are a plurality of sensor terminals 2 and the event generation condition may be changed, the sensor terminal 2 cannot completely determine the event occurrence.

The event occurrence information receiving unit 35 receives event occurrence information including the sampling time from the storage server 1 and sends it to the probability table calculation / holding unit 33.
The priority control unit 36 determines whether or not the event occurrence probability predicted by the event occurrence probability prediction unit 34 exceeds a predetermined occurrence probability. If it is determined that the event occurrence probability exceeds, the sensor information transmission unit 37 is made to perform priority control processing of sensor information. A specific method of this priority control process will be described later.

Next, the operation of the storage server of the sensor information collection system according to the embodiment of the present invention will be described with reference to the flowchart of FIG.
When sensor information is received from the sensor terminal 2 (S1), the received sensor information is stored in the information storage unit 12 (S2). The event detection unit 11 compares the sensor information stored in the information storage unit 12 with a predetermined event generation condition (S3) to determine whether an event needs to be generated (S4). .

  When it is determined that an event needs to be generated, event processing such as mail transmission and alarm generation is performed (S5), and the event information is sent from the event occurrence information feedback unit 13 to the sensor terminal 2 that has transmitted the sensor information. In response to this, event occurrence information is transmitted.

  This event occurrence information includes sampling time (sampling number) and sensor terminal identification information in addition to information indicating that an event has occurred. If it is determined in S4 that it is not necessary to generate an event, or if the process in S6 is terminated, the process is terminated.

Next, the operation of the sensor terminal of the sensor information system according to the embodiment of the present invention will be described with reference to the flowchart of FIG.
First, sensor information is received from the sensor 3 connected to the sensor terminal 2 (S11). Then, processing for generating an event occurrence probability table is performed based on the received sensor information (S12). Thereafter, the event occurrence probability is calculated based on the event occurrence probability table generated in S12 (S13). Based on the calculated event occurrence probability, the priority control process of the sensor information in which the event has occurred is performed. Then, the sensor information on which the priority control process has been performed is transmitted to the storage server 1 (S15), and the process ends.

Hereinafter, the event occurrence probability table generation process in S12, the event occurrence probability calculation process in S13, the priority control process in S14, and the sensor information transmission process to the storage server in S15 will be described in detail.
FIG. 5 is a flowchart for explaining details of the operation of the event occurrence probability table generation processing in S12. The event occurrence probability table generation process is performed by the probability table calculation / holding unit 33.

The probability table calculation / holding unit 33 acquires event occurrence information from the event occurrence information receiving unit 35 (S21), and acquires past sensor information stored in the sensor information temporary storage unit 32 (S22).
Then, a state transition table is created from the acquired past sensor information and event occurrence information (S23). FIG. 6 is a diagram illustrating an example of the relationship between the sensor information table before conversion and the transition table after conversion. As shown in the figure, for each sensor terminal, as shown in FIG. 6, a sensor information table composed of sensor ID, sampling number and sensing data included in sensor information sent from the corresponding sensor, It converts into the state transition table shown in FIG.

The state transition table compares the numerical information of the sensing data with the numerical information of the previous sensing data for each sensor ID, and changes to three state information of increase (↑), no change (→), and decrease (↓). It is converted to simplify the numerical information.
When the state transition table is created in S23, a window for creating a probability table in which the created state transition table is divided into rows and columns is set (S24). In this embodiment, a probability table is created using two samples. As shown in FIG. 8, since the number of sensors is two and the number of samples is two, a total of four tables can be created. Note that the number of samples is determined based on the processing performance of the sensor terminal.

As described above, the reason for the division into rows and columns is that the number of patterns becomes enormous if all the patterns that can be taken by the state transition table are used. For example, in the case of 4 input data and 4 sampling, 3 12 = 531441 pattern.
In the present embodiment, by dividing the state information of the state transition table into rows and columns, the appearance probabilities are calculated, and the number of patterns is reduced using these composite probabilities. In the example of 4 input data and 4 sampling described above, the row has 108 patterns (3 to the power of 3 × 4), the column has 243 patterns (3 to the power of 3 × 3), and a total of 351 patterns. Can be reduced.

  Next, the window set in the state transition transition table in S24 is moved to create a probability table for each window (S25). 9 to 12 show probability tables corresponding to the windows generated by moving the four windows set in the state transition table.

  As shown in FIGS. 9 to 12, the probability table stores the change pattern, the number of event occurrences, the total number of change patterns, and the event occurrence probability. The event occurrence probability is obtained by the number of event occurrences / the total number of change patterns. When the event occurrence probability table is calculated in this way (S12), the event occurrence probability is then calculated (S13).

Hereinafter, a method for calculating the event occurrence probability in S13 will be described. FIG. 13 is a flowchart for explaining the event occurrence probability calculation.
The event occurrence probability predicting unit 34 acquires sensor information from the sensor information receiving unit 31 (S31), converts it into three states, and then uses the event occurrence probability table created in S12 to store each line of the acquired sensor information. And the event occurrence probability of the column is acquired (S32).

Specifically, the event occurrence probability of each window pattern of the acquired sensor information converted into the three states is acquired with reference to the corresponding window event occurrence probability table. FIG. 14 is a diagram for explaining a method for acquiring event occurrence probabilities.
In the figure, when sensor information of sampling numbers “−1” and “0” is acquired, it is converted into a three-state state transition table. And the probability table corresponding to the state of each window set in the acquired sensor information, for example, the window of the row of the sensor 1 is the table 1 shown in FIG. The event occurrence probability “1” corresponding to the window change pattern “→ ↑” is determined. Further, the probability table corresponding to the row window of the sensor 2 is the table 2 shown in FIG. 10, and the event occurrence probability “0” corresponding to the change pattern “↓↓” of the row window of the sensor 2 from the table 2. .5 ". For the columns, event occurrence probabilities “0.25” and “1” are determined by the same method.

  If the event occurrence probability of each row and column of the acquired sensor information is acquired in S32, then the composite probability is calculated from the event occurrence probability of each row and column (S33). This composite probability is calculated by dividing the total event occurrence probability of each window (each row and column) acquired in S32 by the number of windows, that is, the average event occurrence probability of each window as the composite probability. To do.

In the example shown in FIG.
Composite probability = (1 + 0.5 + 1 + 0.25) / 4
= 0.6875
It becomes.
Then, the composite probability calculated in S33 is set as the event occurrence probability (S34), and the event occurrence probability prediction process in S13 in the event occurrence probability prediction unit 34 is ended. After calculating the event occurrence probability in S13, priority control processing of the acquired sensor information is performed in the priority control unit 36 (S14).

FIG. 15 is a flowchart for explaining the priority control process in S14.
The priority control unit 36 acquires the event occurrence probability (composite probability) calculated by the event occurrence probability prediction unit 34 (S41), and determines whether or not the acquired event occurrence probability is higher than a set value. (S42).

  If it is determined in S42 that the value is higher than the set value, the sensor information transmission unit 37 performs an interrupt transmission of the sensor data (S43), the process of the priority control unit 36 is terminated, and it is determined that the value is lower than the set value. If so, the sensor information is added to the transmission queue of the sensor information transmission unit 37 (S44), and the processing of the priority control unit 36 is terminated.

When the priority control process in S14 is completed, next, sensor information is transmitted from the sensor information transmitting unit 37 to the storage server 1 (S15). FIG. 16 is a flowchart for explaining the operation of the sensor information transmission unit 37.
As shown in the figure, the sensor information transmission unit 37 first determines whether or not an interrupt transmission from the priority control unit 36 has occurred (S51). If it is determined in S51 that interrupt transmission has occurred, the interrupt transmission data from the priority control unit 36 is acquired (S52). Then, the acquired interrupt transmission data is transmitted to the storage server 1 (S53), and the processing of the sensor information transmission unit 37 is terminated.

  On the other hand, if it is determined in S51 that no interrupt signal has been generated, the transmission data is acquired from the transmission queue of the sensor information transmission unit 37 (S54), and the acquired interrupt transmission data is transmitted to the storage server 1. (S53), the processing of the sensor information transmission unit 37 is terminated.

  In the above description, the priority control process in the case where the information is interrupted and transmitted when the event occurrence probability exceeds a predetermined set value has been described. However, the present invention is not limited to this. For example, (1) information with high event occurrence probability may be transmitted with priority over other information, or (2) transmission processing of information with low event occurrence probability is interrupted, You may interrupt and transmit, (3) You may change a transmission period with the occurrence probability of an event.

  Therefore, according to the embodiment of the present invention, in a sensor information collection system in which a plurality of sensor terminals and a storage server are connected by a network, each sensor terminal predicts the occurrence probability of an event for the input sensor information. Thus, since priority control of information to be transmitted to the storage server can be performed, it is possible to quickly determine the occurrence of an event in the storage server.

In addition, the learning function using the state transition table can predict sensor information having a high event occurrence probability without setting complicated parameters in the sensor terminal.
Furthermore, by interrupting and transmitting sensor information having a high event occurrence probability to other transmission processes, information having a high event occurrence probability can be quickly transmitted to the storage server even in a low-speed network.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  In addition, the method described in the embodiment is, for example, a magnetic disk (floppy (registered trademark) disk, hard disk, etc.), optical disk (CD-ROM, DVD) as programs (software means) that can be executed by a computer (computer). , MO, etc.), a semiconductor memory (ROM, RAM, flash memory, etc.) or the like, or can be transmitted and distributed via a communication medium. The program stored on the medium side includes a setting program that configures in the computer software means (including not only the execution program but also a table and data structure) to be executed by the computer. A computer that implements this apparatus reads a program recorded on a recording medium, constructs software means by a setting program as the case may be, and executes the above-described processing by controlling the operation by this software means. The recording medium referred to in this specification is not limited to the distribution, but includes a storage medium such as a magnetic disk or a semiconductor memory provided in a computer or a device connected via a network.

1 is an overall view of a sensor information collection system according to an embodiment of the present invention. It is a more specific functional block diagram of a sensor terminal. It is a flowchart for demonstrating operation | movement of the accumulation | storage server of the sensor information collection system which concerns on embodiment of this invention. It is a flowchart for demonstrating operation | movement of the sensor terminal of a sensor information system to embodiment of this invention. It is a flowchart for demonstrating the detail of operation | movement of an event occurrence probability table production | generation process. It is a figure which shows an example of the relationship between the sensor information table before conversion, and the transition table after conversion. It is a figure which shows a state transition table. It is a figure which shows the window for producing a probability table. It is a figure which shows the probability table of the window produced | generated by moving the window set to the state transition table. It is a figure which shows the probability table of the window produced | generated by moving the window set to the state transition table. It is a figure which shows the probability table of the window produced | generated by moving the window set to the state transition table. It is a figure which shows the probability table of the window produced | generated by moving the window set to the state transition table. It is a figure which shows the flowchart for demonstrating event occurrence probability calculation. It is a figure for demonstrating the acquisition method of event occurrence probability. It is a flowchart for demonstrating a priority control process. It is a flowchart for demonstrating operation | movement of a sensor information transmission part.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Accumulation server, 2 ... Sensor terminal, 3 ... Sensor, 11 ... Event detection part, 12 ... Information accumulation part, 13 ... Event generation information feedback part, 21 ... Event prediction part, 22 ... Sensor information transmission part, 31 ... Sensor Information receiving unit, 32 ... sensor information temporary storage unit, 33 ... probability table calculation / holding unit, 34 ... event occurrence probability prediction unit, 35 ... event occurrence information receiving unit, 36 ... priority control unit, 37 ... sensor information transmission unit.

Claims (10)

In a sensor terminal that transmits sensor information to a server that performs event processing based on sensor information,
Receiving means for receiving sensor information including sensor output values from a plurality of sensors;
An event occurrence probability calculating means for calculating an event occurrence probability in the server from the received plurality of sensor information, past sensor information and past event occurrence information;
Priority control means for sending the received plurality of sensor information to the server in preference to other sensor information when the calculated event occurrence probability exceeds a predetermined occurrence probability. Sensor terminal. The event occurrence probability calculating means includes:
Compares past sensor information and the received multiple sensor information with the sensor output value at the most recent sampling time, and samples the status information indicated by increase, no change, or decrease, and past event occurrence information A state transition table for storing time and sensors in association with each other;
A plurality of probability tables that are generated for each of the plurality of windows set in the state transition table, and store the state information pattern in association with the event occurrence probability corresponding to the state information pattern;
2. The sensor terminal according to claim 1, further comprising means for determining an event occurrence probability in the server of the received plurality of sensor information based on the plurality of probability tables.   3. The sensor terminal according to claim 2, wherein the event occurrence probability in the server is an average value of the probabilities obtained by the probability tables.   The sensor terminal according to claim 2, wherein the window is set for each row and column of the state transition table.   The sensor terminal according to claim 1, wherein the event occurrence information is sent from the server.   The sensor terminal according to claim 1, wherein the priority control means transmits the received plurality of pieces of sensor information by interrupt processing. In a sensor information priority control method in a sensor terminal that transmits sensor information to a server that performs event processing based on sensor information,
Receive sensor information including sensor output values from multiple sensors,
From the received plurality of sensor information, past sensor information and past event occurrence information, an event occurrence probability in the server is calculated,
A priority control method, wherein when the calculated event occurrence probability exceeds a predetermined occurrence probability, the received plurality of sensor information is transmitted to the server in preference to other sensor information. The calculation of the occurrence probability of the event is as follows:
Compares past sensor information and the received multiple sensor information with the sensor output value at the most recent sampling time, and samples the status information indicated by increase, no change, or decrease, and past event occurrence information Generate a state transition table to store in association with each time and sensor,
Generated for each of a plurality of windows set in the state transition table, and generates a plurality of probability tables for storing the state information pattern in association with the event occurrence probability corresponding to the state information pattern,
8. The priority control method according to claim 7, wherein an event occurrence probability in the server of the received plurality of sensor information is obtained based on the plurality of probability tables.   9. The priority control method according to claim 8, wherein the event occurrence probability in the server is an average value of the probabilities obtained by the respective probability tables. In a program used in a sensor terminal that transmits sensor information to a server that performs event processing based on sensor information,
The program is stored in the sensor terminal.
Receive sensor information including sensor output values from multiple sensors,
From the received plurality of sensor information, past sensor information and past event occurrence information, the event occurrence probability in the server is calculated,
When the calculated occurrence probability of an event exceeds a predetermined occurrence probability, the program causes the plurality of received sensor information to be transmitted to the server in preference to other sensor information.

Images