KR102480411B1 - Method for Processing Correlationship Edge Computing - Google Patents

Abstract

The present invention relates to a method for processing correlational edge computing, and the method for processing correlational edge computing according to the present invention is a method executed through an edge computer, wherein D (D≥ 1) M (M ≥ 2) sensors to buffer sensing information received from N (N ≥ 2) sensors mounted on or connected to devices and messages received from designated E (E ≥ 1) devices in a first-in-first-out format The first step of creating a queue, W (W≥1) job information to be processed through edge computing, and i (1≤i≤M) to be monitored to process each job for each W job. ), the i-th queue identification information for identifying the monitoring queue, the i-th critical information for the i-th critical condition set in the i-th monitoring queue, and the sensing information or message buffered in the i-th monitoring queue The jth queue identification information identifying the jth (1≤j≤M, i≠j) reference queue to be referred to when the threshold condition of i is exceeded, and the jth threshold set in the jth reference queue. A second step of checking a management table stored in a designated edge storage area by hierarchically including the j-th critical information for a condition, and corresponding to the sensing values of the sensing information received from the N sensors based on the management table. A third step of repeating the process of distributing and buffering data included in messages received from the E number of devices to M queues designated respectively; and i designated for each of the W jobs based on the management table. A fourth step of repeating the process of monitoring whether the specified number of data buffered in the monitoring queue of the i exceeds the critical condition, and the first corresponding to the wth (1≤w≤W) task among the W tasks. When the data buffered in the i monitoring queue deviates from the i-th critical condition, the specified number of data buffered in the j-th reference queue associated with the i-th monitoring queue is transferred to the j-th critical condition based on the management table. A fifth step of checking whether the threshold condition is exceeded, and the data buffered in the i-th monitoring queue escapes the i-th threshold condition and the buffer is stored in the j-th reference queue. When the linked data deviate from the j-th critical condition, the w-th job configured to be processed when the specified number of data buffered in the i-th monitoring queue based on the management table deviates from the i-th critical condition and a sixth step of performing a procedure for processing.

Description

Method for Processing Correlation Edge Computing {Method for Processing Correlationship Edge Computing}

The present invention relates to sensing information received from N (N ≥ 2) sensors mounted on or linked to D (D ≥ 1) devices belonging to the management target of edge computing and designated E (E ≥ 1) devices. Based on M (M ≥ 2) queues buffering messages received from the queue, the specified task processing is performed through correlation verification of the data buffered in each queue.

Edge computing is a concept that contrasts with the concept of sending and processing data generated from various devices to a centralized data center such as the cloud. It is a concept of real-time processing by immediately analyzing without time delay in

On the other hand, for edge computing, a plurality of sensors or devices are connected to one edge computer. Conventional edge computing simply transfers data collected from sensors or devices to an edge computer equipped with an edge service platform through a gateway. (Patent Publication No. 10-2017-0133624 (December 6, 2017)). In this conventional method, the tasks to be processed through the edge computer are simple and the number of tasks is a certain number. On the other hand, if the number of tasks to be processed through the edge computer increases and complexity increases, it has a problem of costly and time-consuming maintenance, especially in the instantaneous (or When a simple malfunction (recoverable) occurs, this simple malfunction affects the entire edge computing.

An object of the present invention to solve the above problems is to respond to sensing information received from N (N ≥ 2) sensors mounted on or linked to D (D ≥ 1) devices belonging to the management target of edge computing. When specific data among data or data included in messages received from at least E (E≥1) devices deviate from the specified critical condition, the specified task corresponding to the critical condition deviation of the specific data is not processed, but the specific data An object of the present invention is to provide a correlation edge computing processing method for processing a specified task corresponding to a critical condition deviation of specific data when the critical condition deviation of other data correlated with the specified threshold condition deviation of data occurs.

The method for processing correlational edge computing according to the present invention is a method executed through an edge computer, which is mounted on or linked to D (D≥1) devices belonging to the management target of edge computing. 2) A first step of creating M (M ≥ 2) queues to buffer sensing information received from sensors and messages received from designated E (E ≥ 1) devices in a first-in-first-out format, and edge W (W ≥ 1) job information to be processed through computing and i-th queue identification information identifying i (1 ≤ i ≤ M) monitoring queues to be monitored in order to process each job for each of the W jobs. and the i-th critical information for the i-th critical condition set in the i-th monitoring queue and the sensing information or message buffered in the i-th monitoring queue deviates from the i-th critical condition. It hierarchically includes the jth queue identification information identifying j (1≤j≤M, i≠j) reference queues and the jth threshold information for the jth threshold condition set in the jth reference queue. A second step of confirming a management table stored in a designated edge storage area, and data corresponding to the sensed values of the sensing information received from the N sensors based on the management table and included in messages received from the E devices A third step of repeating the process of distributing and buffering the buffered data to M queues designated respectively, and the specified number of data buffered in the monitoring queue i designated for each of the W jobs based on the management table is A fourth step of repeating the process of monitoring whether the critical condition of is exceeded, and the data buffered in the i-th monitoring queue corresponding to the w-th task (1≤w≤W) among the W tasks is transferred to the i-th critical condition. a fifth step of checking whether a specified number of data buffered in the j-th reference queue associated with the i-th monitoring queue, based on the management table, deviate from the j-th threshold condition, when the condition is exceeded; When the data buffered in the i-th monitoring queue deviate from the i-th threshold condition and the data buffered in the j-th reference queue deviate from the j-th threshold condition, the management and a sixth step of performing a procedure for processing a w-th job set to be processed when a specified number of data buffered in the i-th monitoring queue deviate from the i-th threshold condition based on the i-th threshold condition.

According to the present invention, in the first step, sensor types for N sensors mounted on or linked to the D devices, sensing value types for sensing values included in sensing information received from the N sensors, and sensors Based on the correlation between at least one of the data types for the data included in the message received from the device type for the D devices mounted, the specified E (E≥1) devices, and each of the W tasks to be processed through edge computing. It may include generating M queues to buffer sensing information or messages received from at least one of N sensors and E devices in a first-in-first-out format.

According to the present invention, the M queues include a queue for buffering sensing information received from designated sensors, a queue for buffering operation information received from designated equipment, and multiple buffering of sensing information and operation information from designated sensors and equipment. Can contain one or more of the queues.

According to the present invention, the M queues multi-buffer the same sensing information received from the same sensor to two or more queues corresponding to monitoring queues of different tasks in order to manage W tasks based on data stored in each queue. It could be possible.

According to the present invention, the E number of devices may be equipped with sensors or overlapping with at least some of the connected D devices.

According to the present invention, the i-th monitoring queue may include one or more queues designated for each task, and the j-th reference queue may include one or more other queues other than the queue included in the i-th monitoring queue. there is.

According to the present invention, the correlational edge computing processing method, when the management table is stored in a designated edge storage area, checks the connection of N sensors connected to the N sensors based on the management table, or connects the E devices A step of confirming the connection of the connected E devices may be further included.

According to the present invention, in the fifth step, when the data buffered in the i-th monitoring queue corresponding to the w-th task deviates from the i-th threshold condition, the procedure for processing the w-th task is not performed. A waiting step may be further included.

According to the present invention, the w-th task recalls the state of a device associated with a sensor corresponding to the sensing information buffered in the i-th monitoring queue or the state of a device corresponding to a message buffered in the i-th monitoring queue. It may include adjusting within the i-th critical condition.

According to the present invention, the management table is stored at the point at which the data buffered in the i-th monitoring queue deviate from the i-th critical condition and the time point at which the data buffered in the j-th reference queue deviated from the j-th critical condition for each task. It further includes time relationship information between points in time, and the data buffered in the M queues further includes a timestamp attached by a sensor or equipment or a timestamp corresponding to a point in time at which the sensing information or message is received, wherein the Step 5 is between the time when the data buffered in the i-th monitoring queue deviated from the i-th threshold condition and the time when the data buffered in the j-th reference queue deviated from the j-th threshold condition based on the timestamp. A step of confirming whether the time relationship of corresponds to the time relationship information of the management table may be further included.

According to the present invention, in the fifth step, when the data buffered in the i-th monitoring queue corresponding to the w-th task does not deviate from the i-th threshold condition, the sensing information buffered in the i-th monitoring queue The method may further include performing a procedure for maintaining a current state of equipment associated with a sensor corresponding to or maintaining a current state of equipment corresponding to a message buffered in the i-th monitoring queue.

According to the present invention, the management table is the kth (1≤k≤M, i≠j≠k) to be referenced when the sensing information or message buffered in the jth reference queue deviates from the jth threshold condition. ) may further include k th queue identification information for identifying the reference queue and k th critical information for a k th critical condition set in the k th reference queue. Meanwhile, in this case, the fifth step further includes confirming that the specified number of data buffered in the k-th reference queue associated with the j-th reference queue deviate from the k-th threshold condition based on the management table. In the sixth step, the data buffered in the i-th monitoring queue deviates from the i-th threshold condition and the data buffered in the j-th reference queue deviates from the j-th threshold condition, and the k-th reference reference queue When the data buffered in the queue deviates from the k-th critical condition, the specified number of data buffered in the i-th monitoring queue based on the management table deviates from the i-th critical condition. It may include performing a procedure for handling the work of w.

According to the present invention, sensing information received from N (N ≥ 2) sensors mounted on or linked to D (D ≥ 1) devices belonging to the management target of edge computing and received from designated E (E ≥ 1) devices Based on M (M ≥ 2) queues buffering messages to be processed, the reliability of each task processing is improved without being controlled or managed through the cloud by performing specified task processing through correlation verification for data buffered in each queue. At the same time, it has the advantage of maximizing the efficiency of management no matter how many tasks.

1 is a diagram showing the configuration of a correlational edge computing system according to an embodiment of the present invention.
2 illustrates a relationship between a monitoring queue and a reference queue according to an embodiment of the present invention.
3 is a diagram illustrating a correlational edge computing process according to an embodiment of the present invention.

Hereinafter, the operating principle of the preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings and description. However, the drawings shown below and the description below relate to preferred implementation methods among various methods for effectively explaining the features of the present invention, and the present invention is not limited to only the drawings and description below.

In addition, in the following description of the present invention, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted. In addition, terms to be described later are terms defined in consideration of functions in the present invention, which may vary according to the intention or custom of a user or operator. Therefore, the definition should be made based on the contents throughout the present invention.

As a result, the technical spirit of the present invention is determined by the claims, and the following embodiments are a means for efficiently explaining the technical spirit of the present invention to those skilled in the art to which the present invention belongs. only

Figure 1 is a diagram showing the configuration of a correlational edge computing system according to an embodiment of the present invention.

In more detail, Figure 1 shows sensing information received from N (N ≥ 2) sensors 140 mounted on or linked to D (D ≥ 1) pieces of equipment 150 belonging to the management target of edge computing. When specific data among data corresponding to or included in messages received from at least E (E≥1) devices 150 deviate from a designated critical condition, a designated task corresponding to the critical condition deviation of the specific data is performed. Rather than processing, when the critical condition deviation of other data correlated with the specified critical condition deviation of the specific data occurs, the configuration of the edge computing system that processes the specified task corresponding to the critical condition deviation of the specific data is shown, Those of ordinary skill in the art to which the present invention pertains may refer to and/or modify this Figure 1 in various implementation methods for the correlated edge computing system (eg, some components are omitted, subdivided, or Combined implementation methods) may be inferred, but the present invention includes all the implementation methods inferred above, and its technical characteristics are not limited only to the implementation methods shown in FIG. 1.

The system of the present invention transmits messages to N (N ≥ 2) sensors 140 and edge computers 100 mounted on or linked to D (D ≥ 1) devices 150 belonging to the management target of edge computing. Sensing information is received from E (E≥1) devices 150 that exchange and the N sensors 140 through a designated communication interface or a specified message is received from the E devices 150, and a designated message is received based on this. It includes an edge computer 100 (Edge Computer) that processes tasks and is connected to the cloud through a communication network.

The N sensors 140 are mounted on the equipment 150 belonging to the management target of edge computing to sense the state of the equipment 150 or to sense the state associated with the equipment 150 in connection with the equipment 150 As a generic name for the device, for example, a temperature sensor 140, a pressure sensor 140, an optical sensor 140, an acoustic sensor 140, a magnetic sensor 140, a radiation sensor 140, a chemical sensor 140, It may include various sensors 140 for sensing the state of the device 150, such as the biosensor 140, or the state associated with the device 150. The N sensors 140 sense the state of the equipment 150 or the state associated with the equipment 150 in the form of electrical signals, generate sensing information including a sensing value in the form of digitized data, and designate a communication interface. The sensed sensing information may be transmitted to the designated edge computer 100 through

According to the implementation method of the present invention, the N sensors 140 each sense a state according to a designated sensing cycle, and the state sensing cycle for each sensor 140 may be different from each other. Meanwhile, as an embodiment of the present invention, the sensor 140 may attach a timer to count the time synchronized with a designated reference time (eg, GMT standard time or the internal time of the edge computer 100). In this case, The sensor 140 may attach a timestamp corresponding to the time counted by the timer to the sensing information, but the present invention is not limited thereby.

The E number of devices 150 are devices 150 capable of transmitting a message including specified data to the edge computer 100 through a designated communication interface among the devices 150 belonging to the management target of edge computing, and the E number of devices 150 At least some of the devices 150 of 150 may receive messages transmitted from the edge computer 100 .

According to the implementation method of the present invention, the E number of devices 150 may include an intersection set overlapping with at least some of the D devices 150 equipped with or associated with the sensor 140. Alternatively, the E number of devices 150 may be a subset of the D number of devices 150 equipped with or associated with the sensor 140 . For example, the equipment 150 belonging to the edge computing target may include a Programmable Logic Controller (PLC).

Meanwhile, it is preferable that D devices 150 or E devices 150 belonging to the edge computing target of a specific edge computer 100 do not overlap with other edge computers 100 and edge computing targets and maintain independence. Preferably, one edge computer 100 and D devices 150 or E devices 150 belonging to the edge computing target may form one independent ecosystem that interworks in real time. However, the edge computer 100 may interwork with a terminal (eg, a wireless terminal, a wired terminal, etc.) or a server in the cloud through a communication network in non-real time.

Referring to Figure 1, the edge computer 100 receives information from N (N ≥ 2) sensors 140 mounted on or connected to D (D ≥ 1) devices 150 belonging to the management target of edge computing. A queue generator 105 for generating M (M ≥ 2) queues to buffer sensing information and messages received from designated E (E ≥ 1) devices 150 in a first-in-first-out format; W job information to be processed through computing and the ith queue identification information identifying the ith (1≤i≤M) monitoring queue to be monitored for each of the W jobs and the ith The i-th critical information for the i-th critical condition set in the monitoring queue and the j-th (1≤j ≤ M, i ≠ j) designates a management table hierarchically including the j-th queue identification information identifying the reference queue and the j-th threshold information for the j-th critical condition set in the j-th reference queue A table management unit 110 that stores in an edge storage area or checks a management table stored in the designated edge storage area, and data corresponding to the sensing values of the sensing information received from the N sensors 140 based on the management table. and a buffering processing unit 115 repeating a process of distributing and buffering data included in messages received from the E number of devices 150 to each designated M queue, and for each of the W tasks based on the management table The queue monitoring unit 120 repeats a process of monitoring whether a specified number of data buffered in the specified i-th monitoring queue deviates from the i-th critical condition, and the wth (1≤w≤W) of the W tasks If the data buffered in the i-th monitoring queue corresponding to the task of deviate from the i-th critical condition, the specified number of data buffered in the j-th reference queue associated with the i-th monitoring queue based on the management table. A queue referrer 125 for checking whether the data exceeds the j-th threshold condition, and the j-th reference reference unit 125 when the data buffered in the i-th monitoring queue deviates from the i-th threshold condition. When the data buffered in the queue deviates from the j-th critical condition, the specified number of data buffered in the i-th monitoring queue based on the management table deviates from the i-th critical condition. It functionally implements the task processing unit 130 that performs the procedure for processing the task, and may include a cloud interworking unit 135 that interworks with a terminal or server on the cloud in non-real time through a communication network.

The queue generator 105 senses sensing information through N sensors 140 mounted on or connected to D devices 150 among the devices 150 belonging to the management target of edge computing and received through a designated communication interface. M number of buffers for buffering and/or buffering messages generated according to a specified procedure and received through a specified communication interface by events occurring in E specified devices 150 among the devices 150 belonging to the management target of edge computing. create a queue

According to the implementation method of the present invention, the queue generator 105 includes sensor types, N sensors ( 140), the type of sensing value for the sensing value included in the sensing information received, the device type for the D devices 150 equipped with the sensors 140, and included in the messages received from the designated E devices 150 Sensing information or messages received from one or more of the N sensors 140 and E devices 150 based on the correlation between one or more of the data types for the processed data and each of the W tasks to be processed through edge computing You can create M queues to buffer.

According to the implementation method of the present invention, the queue generator 105 is configured to buffer sensing information received from N sensors 140 and messages received from designated E equipment 150 according to a programmed procedure. , and the procedure of generating the M queues by the queue creation unit 105 may be set or programmed by interworking with a server or terminal in the cloud through the cloud interworking unit 135 .

According to the implementation method of the present invention, the M queues include a queue for buffering sensing information received from a designated sensor 140, a queue for buffering operation information received from a designated equipment 150, and a designated sensor 140. It may include one or more queues among queues for multi-buffering sensing information and operating information from the equipment 150.

According to the implementation method of the present invention, the M queues correspond to monitoring queues of different tasks by using the same sensing information received from the same sensor 140 to manage W tasks based on data stored in each queue. You can multi-buffer to more than one queue. That is, in the present invention, the M queues are created to monitor and process tasks to be processed in units of queues to control and manage through edge computing, and are different from well-known queues created for each data to be buffered.

The table management unit 110 includes W job information defined to be processed using one or more devices 150 belonging to the management target of edge computing, and needs to be monitored to process each job for each of the W jobs. The i-th queue identification information for identifying the i-th (1≤i≤M) monitoring queue, the i-th critical information for the i-th critical condition set in the i-th monitoring queue, and the i-th monitoring queue The jth queue identification information identifying the jth (1≤j≤M, i≠j) reference queue to be referred to when the buffered sensing information or message deviates from the ith threshold condition and the jth queue identification information. A management table hierarchically containing the j-th threshold information for the j-th threshold condition set in the reference queue is provided in the edge computer 100 or stored in a designated edge storage area linked to the edge computer 100, or A management table recently stored (or updated) in a designated edge storage area may be checked and provided to the queue monitoring unit 120, the queue reference unit 125, and the job processing unit 130.

Meanwhile, the i-th monitoring queue to be monitored to process each task includes one or more queues designated for each task, and when the sensing information or message buffered in the i-th monitoring queue exceeds the i-th critical condition. The j-th reference queue to be referred to may include one or more other queues other than the queue included in the i-th monitoring queue.

According to the implementation method of the present invention, the management table stored in the edge storage area may be set by a server or terminal in the cloud through the cloud linking unit 135 or downloaded from a server or terminal in the cloud.

Meanwhile, according to the implementation method of the present invention, the management table determines when the data buffered in the i-th monitoring queue deviates from the i-th threshold condition for each task and when the data buffered in the j-th reference queue deviates from the j-th critical condition. Time relation information between time points out of conditions may be further included, and the present invention is not limited thereby. On the other hand, when the management table includes the time relationship information, it is preferable that the M queues are generated in a size capable of discriminating the time relationship information based on the sensing period for each N number of sensors 140 .

Meanwhile, according to the implementation method of the present invention, the management table is the kth (1≤k≤M, i k th queue identification information for identifying the reference queue ≠ j ≠ k) and k th critical information about the k th critical condition set in the k th reference queue. Meanwhile, the hierarchy of reference queues can be additionally extended from the j-th reference queue and the k-th reference queue, and the present invention is not limited thereby.

When the management table is stored in the designated edge storage area, the buffering processor 115 checks the connection of the N sensors connected to the N sensors 140 or connects the E devices 150 based on the management table. E device connections are confirmed, and based on the management table, the data corresponding to the sensing values of the sensing information received from the N sensors 140 and the data included in the messages received from the E devices 150 are The process of distributing and buffering to each designated M queue is repeated.

According to the implementation method of the present invention, the buffering processing unit 115 can multi-buffer the same sensing information or message received from the same sensor 140 or equipment 150 in different queues based on the management table, , The present invention is not limited thereby.

According to the implementation method of the present invention, the time when the data buffered in the i-th monitoring queue for each task in the management table deviates from the i-th critical condition and the j-th critical condition when the data buffered in the j-th reference queue deviates from the i-th critical condition. If the time relationship information between the time points of departure is further included, the data corresponding to the sensing values of the sensing information buffered in the M queues or the data corresponding to the message is sent to each sensor 140 or equipment 150. A timestamp corresponding to the time counted by the provided timer may be attached, or a timestamp corresponding to the internal time of the edge computer 100 at the time when the buffering processing unit 115 receives the sensing information or message may further include.

The data corresponding to the sensing values of the sensing information received from the N sensors 140 and the data included in the messages received from the E devices 150 are distributed and buffered in M queues through the buffering processor 115. If it is, the queue monitoring unit 120 repeats the process of monitoring whether the specified number of data buffered in the i-th monitoring queue for each of the W tasks deviates from the i-th threshold condition based on the management table. do.

If the data buffered in the i-th monitoring queue corresponding to the w-th task among the W tasks does not exceed the i-th threshold condition, the task processing unit 130 determines the i-th monitoring queue buffered in the i-th monitoring queue. The current state of the equipment 150 associated with the sensor 140 corresponding to the sensing information is maintained, or the current state of the equipment 150 corresponding to the message buffered in the i-th monitoring queue is not performed by a separate procedure, Alternatively, the current state of the device 150 associated with the sensor 140 corresponding to the sensing information buffered in the i-th monitoring queue is maintained or the current state of the device 150 corresponding to the message buffered in the i-th monitoring queue is maintained. You can perform established procedures to maintain the state.

Meanwhile, if the data buffered in the i-th monitoring queue corresponding to the w-th task among the W tasks deviates from the i-th threshold condition, the task processing unit 130 performs a procedure for processing the w-th task. It is preferable to wait until it is confirmed that the specified number of data buffered in the j-th reference queue associated with the i-th monitoring queue through the queue reference unit 125 deviate from the j-th threshold condition, rather than immediately. desirable. For example, the queue monitoring unit 120 buffers the sensing information sensed by a specific fire sensor (= chemical sensor) 140 associated with the equipment 150 belonging to the management target of edge computing. Check whether the data buffered in deviate from the i-th critical condition, and if the data buffered in the i-th queue for buffering the sensing information sensed by the specific fire sensor 140 deviate from the i-th critical condition Even if it is sensed that a fire has occurred, the task processing unit 130 may stand by without immediately issuing a fire alarm.

If the data buffered in the i-th monitoring queue corresponding to the w-th task deviates from the i-th critical condition, the queue reference unit 125 determines the i-th monitoring queue and the i-th monitoring queue based on the management table. A procedure for checking whether the specified number of data buffered in the associated j-th reference queue deviates from the j-th threshold condition is performed. For example, the queue monitoring unit 120 buffers the sensing information sensed by the specific fire sensor 140 associated with the equipment 150 belonging to the management target of edge computing. It is checked whether the i-th critical condition is exceeded, and if the data buffered in the i-th queue that buffers the sensing information sensed by the specific fire sensor 140 deviate from the i-th critical condition, it is sensed that a fire has occurred. In this case, the queue reference unit 125 determines that the data buffered in the j-th reference queue buffering the sensing information sensed by the specific optical sensor 140 associated with the equipment 150 meets the j-th threshold condition. It is possible to determine whether a luminous intensity change corresponding to a spark or flame corresponding to a fire has been sensed. Here, since the change in luminous intensity can be sensed even by a light, it is preferable to set it as a reference object rather than a monitoring object for a fire alarm task.

On the other hand, according to the extended implementation method of the present invention, the data buffered in the i-th monitoring queue for each task in the management table deviates from the i-th critical condition and the data buffered in the j-th reference queue Time relationship information between times when the critical condition is deviated is included, and a timestamp attached by the sensor 140 or device 150 to the data buffered in the M queues or the time when the sensing information or message is received When the corresponding timestamp is included, the queue reference unit 125 determines the point at which the data buffered in the i-th monitoring queue deviated from the i-th threshold condition and the buffered data in the j-th reference queue based on the timestamp. A process of confirming whether the time relationship between the points at which the updated data deviates from the j-th critical condition corresponds to the time relationship information of the management table may be further performed. For example, whether sparks or flames corresponding to the fire were sensed through the optical sensor 140 within 3 minutes before sensing the fire through the fire sensor 140 in the management table corresponding to the above fire alarm example. When time relationship information for confirmation is included, the queue reference unit 125 selects data corresponding to data within 3 minutes before a fire is sensed through the i-th monitoring queue among a plurality of data buffered in the j-th reference queue. Among them, it may be confirmed whether a spark or a flame corresponding to a fire is sensed because the specified number of data deviate from the j-th critical condition.

Meanwhile, it is confirmed through the queue monitoring unit 120 that the data buffered in the i-th monitoring queue deviates from the i-th critical condition, and is associated with the i-th monitoring queue through the queue reference unit 125. When it is confirmed that the data buffered in the j-th reference queue deviates from the j-th threshold condition, the task processing unit 130 determines the specified number of data buffered in the i-th monitoring queue based on the management table. A procedure for processing the w-th job set to be processed when i deviated from the i-th critical condition can be performed. For example, data buffered in the i-th queue that buffers sensing information sensed by a specific fire sensor 140 associated with the equipment 150 belonging to the management target of edge computing through the queue monitoring unit 120 is out of the i-th critical condition, and the occurrence of a fire is confirmed, and the j-th buffering the sensing information sensed by the specific optical sensor 140 associated with the equipment 150 through the cue reference unit 125. When it is confirmed that the data buffered in the reference queue deviate from the j-th threshold condition and the change in light intensity corresponding to a spark or flame corresponding to a fire is sensed, the task processing unit 130 senses by the fire sensor 140 When data buffered in the i-th queue for buffering the detected sensing information deviates from the i-th threshold condition, a procedure for processing the w-th task corresponding to the issuance of a fire alarm set to be processed may be performed. On the other hand, the present invention improves the reliability of each job process by performing a specified job process through correlation verification for data buffered in each queue based on M queues, and at the same time, it is possible to maximize the efficiency of management no matter how many jobs there are. .

According to the implementation method of the present invention, the w-th task is the state of the equipment 150 associated with the sensor 140 corresponding to the sensing information buffered in the i-th monitoring queue or the buffered in the i-th monitoring queue. It may include adjusting the state of the device 150 corresponding to the message to within the i-th threshold condition.

Meanwhile, according to the extended implementation method of the present invention, the kth (1≤k≤M, When the kth queue identification information identifying the reference queue i≠j≠k) and the kth threshold information for the kth critical condition set in the kth reference queue are further included, the queue reference unit ( 125) may further perform a process of confirming that a specified number of data buffered in the k-th reference queue associated with the j-th reference queue deviate from the k-th threshold condition based on the management table. In this case, the task processing unit 130 determines that the data buffered in the i-th monitoring queue deviate from the i-th threshold condition, the data buffered in the j-th reference queue deviate from the j-th threshold condition, and the k-th threshold condition. When the data buffered in the reference queue deviates from the k-th critical condition, the specified number of data buffered in the i-th monitoring queue based on the management table deviates from the i-th critical condition. The procedure for handling the wth job can be performed. For example, if the management table corresponding to the above fire alarm example is set to refer to the temperature sensor 140 in addition to the optical sensor 140, the cue reference unit 125 is the equipment 150 ) The data buffered in the j-th reference queue that buffers the sensing information sensed by the specific optical sensor 140 associated with ) deviates from the j-th critical condition, resulting in a change in light intensity corresponding to a spark or flame corresponding to a fire. At the same time (or after) that the sensing is confirmed, the data buffered in the kth reference queue that buffers the sensing information sensed by the specific temperature sensor 140 associated with the equipment 150 satisfies the kth critical condition. It is possible to check whether a temperature rise corresponding to a fire is sensed by leaving, and after the occurrence of a fire is sensed through the queue monitoring unit 120, the task processing unit 130 changes light intensity and temperature through the cue reference unit 125. When all rises are sensed, a fire alarm procedure corresponding to the fire detection may be performed.

Figure 2 illustrates a relationship between a monitoring queue and a reference queue according to an embodiment of the present invention.

In more detail, Figure 2 shows the i (1 ≤ i ≤ M) number of tasks to be monitored in order to process the w (1 ≤ w ≤ W) task among W (W ≥ 1) tasks to be processed through edge computing. The relationship between the watch queue and the jth (1≤j≤M, i≠j) reference queue to be referred to when the data buffered in the i-th watchdog deviates from the i-th threshold condition and the i-th reference queue It is a diagram illustrating a relationship between data that deviate from the i-th threshold condition among data buffered in the monitoring queue of and data that deviate from the j-th threshold condition among data buffered in the j-th reference queue. Those with ordinary knowledge in the technical field to which the present invention belongs will be able to infer various extended embodiments, such as an embodiment in which the reference queue is extended by referring to and/or modifying Figure 2, and the present invention includes all examples inferred above.

Referring to the example of FIG. 2, the i-th monitoring queue and the j-th reference queue are separate queues that are distinguished from each other, and among the data buffered in the i-th monitoring queue, data that deviate from the i-th threshold condition and Among the data buffered in the j-th reference queue, a time difference within the preset time relationship information may occur between data that deviate from the j-th threshold condition.

Figure 3 is a diagram illustrating a correlational edge computing process according to an embodiment of the present invention.

In more detail, FIG. 3 shows sensing information received from N (N ≥ 2) sensors 140 mounted on or linked to D (D ≥ 1) pieces of equipment 150 belonging to the management target of edge computing and a designated E ( Based on M (M ≥ 2) queues buffering messages received from E≥1) number of devices 150, a process of performing a specified task process through correlation verification for data buffered in each queue is shown. , Those of ordinary skill in the art to which the present invention pertains, various implementation methods for the process (eg, some steps are omitted or the order is changed) by referring to and / or modifying this Figure 3 Although it may be inferred, the present invention is made including all the implementation methods inferred above, and its technical characteristics are not limited only to the implementation method shown in FIG. 3.

Referring to Figure 3, the edge computer 100 has W (W≥1) individual job information to be processed through edge computing and i (1≤i≤) to be monitored to process each job for each W job. The i-th queue identification information for identifying the monitoring queue of M), the i-th critical information for the i-th critical condition set in the i-th monitoring queue, and the sensing information or message buffered in the i-th monitoring queue The jth queue identification information identifying the jth (1≤j≤M, i≠j) reference queue to be referred to when the ith threshold condition is exceeded, and the jth queue set in the jth reference queue. The management table hierarchically including the j-th critical information for the critical condition is stored in the designated edge storage area (300). Meanwhile, the management table indicates the time between the time when the data buffered in the i-th monitoring queue deviate from the i-th critical condition and the time when the data buffered in the j-th reference queue deviated from the j-th critical condition for each task. The kth (1≤k≤M, i≠j≠k) to be referenced when the sensing information or message further including relationship information and/or buffered in the jth reference queue deviates from the jth threshold condition. ) may further include k th queue identification information for identifying the reference queue and k th critical information for a k th critical condition set in the k th reference queue.

Based on the management table, the edge computer 100 controls sensor types for the N sensors 140 mounted on or connected to the D devices 150 and sensing information received from the N sensors 140. Sensed value type for included sensing values, device type for D devices 150 equipped with sensors 140, data included in messages received from designated E (E≥1) devices 150 Sensing information or messages received from one or more of the N sensors 140 and E devices 150 are first-in-first-out based on the correlation between one or more of the data types and each of the W tasks to be processed through edge computing. M queues to be buffered are created (305).

If the M queues are created, the edge computer 100 checks the management table stored in the designated edge storage area (310), and communicates with the N sensors 140 through a communication interface based on the management table. It checks the connection of the connected N sensors or checks the connection of the E devices connected to the E devices 150 (315). If the N sensor connection and the E device connection are not confirmed, the edge computer 100 may perform a procedure for communicating with the sensor 140 or device 150 that is not connected through the communication interface. (320).

Meanwhile, when the connection of the N sensors and the connection of the E devices are confirmed, the edge computer 100 stores data corresponding to the sensing values of the sensing information received from the N sensors 140 based on the management table and the E While repeating the process of distributing and buffering the data included in the messages received from the number of devices 150 to each designated M queue (325), the monitoring queue i designated for each of the W jobs based on the management table. The process of monitoring whether the specified number of data buffered in deviate from the i-th threshold condition is repeated (330).

If the data buffered in the i-th monitoring queue corresponding to the w-th task does not deviate from the i-th critical condition, the edge computer 100 responds to the sensing information buffered in the i-th monitoring queue. The current state of the equipment 150 associated with the sensor 140 to be maintained or the current state of the equipment 150 corresponding to the message buffered in the i-th monitoring queue is maintained (335).

Meanwhile, if the data buffered in the i-th monitoring queue corresponding to the w-th task (1≤w≤W) out of the W tasks deviates from the i-th critical condition, the edge computer 100 operates the management table. Based on , it is checked whether the specified number of data buffered in the j-th reference queue associated with the i-th monitoring queue deviate from the j-th threshold condition (340). On the other hand, in the management table, the time between the time when the data buffered in the i-th monitoring queue for each job deviated from the i-th critical condition and the time when the data buffered in the j-th reference queue deviated from the j-th critical condition When the relation information is further included, the edge computer 100 determines the point at which the data buffered in the i-th monitoring queue deviate from the i-th threshold condition based on the timestamp included in the data buffered in the M queues. A process of confirming whether a time relationship between the times when data buffered in the j-th reference queue deviate from the j-th threshold condition may correspond to the time relationship information of the management table may be further performed.

If the data buffered in the i-th monitoring queue deviate from the i-th threshold condition, but the data buffered in the j-th reference queue does not deviate from the j-th threshold condition, the edge computer 100 Maintains the current state of the device 150 associated with the sensor 140 corresponding to the sensing information buffered in the i-th monitoring queue or displays the current state of the device 150 corresponding to the message buffered in the i-th monitoring queue. Can be maintained(335).

Meanwhile, when the data buffered in the i-th monitoring queue deviate from the i-th threshold condition and the data buffered in the j-th reference queue deviate from the j-th threshold condition, the edge computer 100 Based on the management table, when the specified number of data buffered in the i-th monitoring queue deviate from the i-th threshold condition, a procedure for processing the w-th job set to be processed is performed (345). Meanwhile, in the management table, the k-th (1≤k≤M, i≠j≠k) reference queue to be referenced when the sensing information or message buffered in the j-th reference queue deviates from the j-th threshold condition. If the kth queue identification information identifying the kth queue identification information and the kth critical information for the kth critical condition set in the kth reference queue are further included, the edge computer 100 based on the management table A process of confirming that the specified number of data buffered in the k-th reference queue associated with the j-th reference queue deviate from the k-th threshold condition may be further performed. In this case, the edge computer 100 When the data buffered in the i-th monitoring queue deviates from the i-th threshold condition, the data buffered in the j-th reference queue deviates from the j-th threshold condition, and the data buffered in the k-th reference queue deviates from the i-th threshold condition. A procedure for processing the w-th job set to be processed when the specified number of data buffered in the i-th monitoring queue based on the management table deviate from the i-th critical condition when the k critical condition is exceeded can be performed.

Meanwhile, when the wth job set to be processed when the specified number of data buffered in the ith monitoring queue deviates from the ith critical condition, the edge computer 100 determines based on the management table The process of distributing and buffering the data corresponding to the sensing values of the sensing information received from the N sensors 140 and the data included in the messages received from the E devices 150 to each designated M queue is repeated, or And/or it can be repeated from the process of confirming N sensor connections or E device connections, and it is preferable to repeat until the edge computer 100 of FIG. 3 is terminated.

100: edge computer 105: queue generation unit
110: table management unit 115: buffering processing unit
120: queue monitoring unit 125: queue reference unit
130: work processing unit 135: cloud interlocking unit
140: sensor 150: equipment

Claims (13)

In the method executed through the edge computer,
Sensing information received from N (N ≥ 2) sensors mounted on or linked to D (D ≥ 1) devices belonging to the management target of edge computing and messages received from designated E (E ≥ 1) devices are first-in-first-out. A first step of generating M (M≥2) queues to be buffered in the form;
Identification of the i-th queue that identifies W (W≥1) job information to be processed through edge computing and i-th (1≤i≤M) monitoring queues to be monitored to process each job for each of the W jobs information and the i-th critical information for the i-th critical condition set in the i-th monitoring queue and the sensing information or message buffered in the i-th monitoring queue deviates from the i-th critical condition. The j-th queue identification information for identifying the j-th reference queue (1≤j≤M, i≠j) and the j-th threshold information for the j-th threshold condition set in the j-th reference queue are hierarchically A second step of checking the management table stored in the designated edge storage area including;
Based on the management table, the data corresponding to the sensing values of the sensing information received from the N sensors and the data included in the messages received from the E devices are distributed to designated M queues and buffered. Repeating the process third step;
a fourth step of repeating a process of monitoring whether a specified number of data buffered in the i-th monitoring queue for each of the W tasks deviates from the i-th threshold condition based on the management table;
If the data buffered in the i-th monitoring queue corresponding to the w-th task (1≤w≤W) out of the W tasks deviates from the i-th critical condition, the i-th monitoring queue is based on the management table. a fifth step of checking whether the specified number of data buffered in the j-th reference queue associated with the queue deviates from the j-th threshold condition; and
When the data buffered in the i-th monitoring queue deviate from the i-th threshold condition and the data buffered in the j-th reference queue deviate from the j-th threshold condition, the th threshold condition is exceeded based on the management table. A sixth step of performing a procedure for processing the wth job set to be processed when the specified number of data buffered in i's monitoring queue deviate from the ith critical condition.
The method of claim 1, wherein the first step,
Sensor types for the N sensors mounted on or linked to the D devices, sensing value types for sensing values included in sensing information received from the N sensors, equipment types for the D devices equipped with sensors, One of N sensors and E devices based on the correlation between one or more of the data types for data contained in messages received from the specified E (E≥1) devices and each of the W tasks to be processed through edge computing. A correlational edge computing processing method comprising the step of generating M queues to buffer sensing information or messages received from the above in a first-in-first-out format.
The method of claim 1, wherein the M queues,
A queue for buffering sensing information received from a designated sensor;
A queue for buffering operation information received from designated equipment;
A correlational edge computing processing method comprising at least one of queues for multi-buffering sensing information and operation information from designated sensors and equipment.
The method of claim 1 or 3, wherein the M queues,
In order to manage W tasks based on the data stored in each queue, the same sensing information received from the same sensor can be multi-buffered in two or more queues corresponding to the monitoring queues of each different task. Correlational edge computing processing. method.
The method of claim 1, wherein the E pieces of equipment,
Correlated edge computing processing method, characterized in that the sensor is mounted or overlapped with at least some of the D equipment associated with it.
According to claim 1,
The i-th monitoring queue includes one or more queues designated for each task,
The j-th reference queue includes one or more other queues other than the queue included in the i-th monitoring queue.
According to claim 1,
If the management table is stored in the designated edge storage area,
Correlational edge computing processing method, characterized in that further comprising the step of checking the connection of N sensors connected to the N sensors or checking the connection of E devices connected to the E devices based on the management table.
The method of claim 1, wherein the fifth step,
If the data buffered in the i-th monitoring queue corresponding to the w-th task deviate from the i-th critical condition, waiting without performing a procedure for processing the w-th task Characterized by a correlational edge computing processing method.
The method of claim 8, wherein the wth operation,
Adjusting a state of a device associated with a sensor corresponding to sensing information buffered in the i-th monitoring queue or a state of a device corresponding to a message buffered in the i-th monitoring queue to within the i-th threshold condition Correlated edge computing processing method, characterized in that made by.
According to claim 1,
The management table indicates the time between the time when the data buffered in the i-th monitoring queue deviated from the i-th critical condition and the time when the data buffered in the j-th reference queue deviated from the j-th critical condition for each task. Including more relational information,
The data buffered in the M queues further includes a timestamp attached by a sensor or equipment or a timestamp corresponding to a time point at which the sensing information or message is received,
The fifth step,
Based on the timestamp, the time relationship between the time when the data buffered in the i-th monitoring queue deviated from the i-th critical condition and the time when the data buffered in the j-th reference queue deviated from the j-th critical condition was Correlated edge computing processing method characterized in that it further comprises the step of checking whether it corresponds to the time relationship information of the management table.
The method of claim 1, wherein the fifth step,
If the data buffered in the i-th monitoring queue corresponding to the w-th task does not deviate from the i-th critical condition, the current state of the equipment associated with the sensor corresponding to the sensing information buffered in the i-th monitoring queue or performing a procedure for maintaining a current state of a device corresponding to a message buffered in the i-th monitoring queue.
The method of claim 1, wherein the management table,
The kth reference queue identifying the kth (1≤k≤M, i≠j≠k) reference queue to be referenced when the sensing information or message buffered in the jth reference queue exceeds the jth threshold condition. and kth critical information for the kth critical condition set in the kth reference queue and the queue identification information of the kth reference queue.
According to claim 12,
The fifth step further includes confirming that a specified number of data buffered in a k-th reference queue associated with the j-th reference queue deviate from a k-th threshold condition based on the management table,
The sixth step,
When the data buffered in the i-th monitoring queue deviates from the i-th threshold condition, the data buffered in the j-th reference queue deviates from the j-th threshold condition, and the data buffered in the k-th reference queue deviates from the i-th threshold condition. When the critical condition of k is exceeded, a procedure for processing the wth job set to be processed when the specified number of data buffered in the monitoring queue of the ith based on the management table deviate from the ith critical condition Correlated edge computing processing method comprising the step of performing.

Images