An oil and gas internet of things energy management method based on data fusion

By optimizing the working mode and data filtering of sensor nodes through a data fusion-based energy management method, the problem of low energy management efficiency of sensor nodes in oil and gas transportation pipelines is solved, and efficient utilization of network energy consumption and fault prevention are achieved.

CN116822824BActive Publication Date: 2026-11-10WUXI UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310180092.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-28
Publication Date
2026-11-10
Estimated Expiration
2043-08-28

AI Technical Summary

Technical Problem

The limited wireless transmission and computing capabilities of sensor nodes in oil and gas pipelines, coupled with the difficulty in replacing batteries due to their deployment in harsh environments, result in low network energy management efficiency and increased risk of failure.

Method used

An energy management method based on data fusion is adopted, which optimizes the working mode and data filtering of sensor nodes by dividing the cluster and selecting the cluster head, thereby reducing redundant information transmission and improving network energy efficiency.

Benefits of technology

It effectively reduces the communication distance between sensor nodes and the aggregation node, avoids redundant information consumption, improves network energy efficiency, and extends network lifespan.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116822824B_ABST
    Figure CN116822824B_ABST
Patent Text Reader

Abstract

The application discloses an energy management method based on data fusion and belongs to the field of oil and gas Internet of Things. Under the background of the oil and gas Internet of Things, the application realizes the prolongation of the service life of the oil and gas Internet of Things and ensures the communication quality of data through the optimization selection of a cluster head based on a core area and an energy management algorithm based on data fusion.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of oil and gas Internet of Things (IoT) technology, specifically to an oil and gas IoT energy management method based on data fusion. Background Technology

[0002] Oil and gas pipelines are typically deployed in sparsely populated or harsh natural conditions, making manual monitoring and inspection extremely difficult. Oil and gas transportation involves long-distance pipeline transport, increasing the likelihood of pipeline failures. Once a pipeline malfunctions, such as leaks or abnormal temperatures, occur, it will seriously threaten the safety of residents and property in the surrounding area, as well as the ecological environment. Deploying sensor nodes at regular intervals between the first, intermediate, and last stations of the pipeline allows for real-time monitoring of various data points during pipeline transportation, such as pressure and temperature. Managers can monitor the status of the oil and gas pipeline in real time, effectively preventing pipeline failures and reducing pipeline maintenance costs.

[0003] Currently, sensors possess wireless transmission and reception capabilities, but their storage and computing power are limited, and they are battery-powered. Therefore, the limited resources of each node restrict the performance of wireless sensor networks. Energy is one of the most significant factors limiting network performance. When sensor nodes are deployed in harsh environments inaccessible to humans, replacing batteries on a large scale is an impossible task. Therefore, effective energy management of sensor networks is the biggest challenge in optimizing the performance of the oil and gas Internet of Things (IoT).

[0004] Therefore, based on the above problems, we propose an oil and gas Internet of Things (IoT) energy management method based on data fusion. This method can reduce the communication distance between sensor nodes and the aggregation node, improve network energy efficiency, reduce or avoid redundant information consuming a lot of energy, and determine the degree of data anomaly to avoid duplication of network information and excessive energy consumption. This is very necessary. Summary of the Invention

[0005] The purpose of this invention is to provide an oil and gas Internet of Things energy management method based on data fusion to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an oil and gas Internet of Things energy management method based on data fusion, comprising the following steps:

[0007] S1: Divide the core region and the edge region based on the size of the cluster. Select the cluster head in the core region by calculation. In this process, all nodes in the network need to calculate and store their remaining energy. Select the node with the most remaining energy in the core region as the cluster head.

[0008] S2: After selecting the cluster head, calculate the cluster head term based on the remaining energy;

[0009] S3: Nodes in the cluster that are not selected as cluster heads calculate their working modes based on their remaining energy, and classify them into working modes: working period, listening period, and dormant period.

[0010] During the working period, the node is in normal working condition, completing tasks such as energy harvesting, environmental monitoring, spectrum sensing, and data uploading;

[0011] During the listening period, the node is responsible for spectrum sensing, environmental monitoring, and energy harvesting. If it senses that the spectrum is idle, the node will also transmit data.

[0012] During the dormant period, the node is only responsible for energy collection;

[0013] S4: Data in the cluster that is not selected as the cluster head is filtered according to the data filtering principle. Only important data is selected for uploading. The data filtering is divided according to the degree of deviation of the data. The greater the degree of deviation, the more abnormal the detection environment is.

[0014] S5: Data filtering and fusion between clusters.

[0015] Preferably, the specific formula for calculating the remaining energy of the node in step S1 is as follows:

[0016] E th,res >ω×E CH,i ,(ω=3);

[0017] E th,CH =E th,res ;

[0018] Where CH represents the cluster head, E th,res E represents the remaining energy of CH. CH,i E represents the energy consumed by a CH node during one normal operation. th,CH This represents the energy threshold of CH.

[0019] Preferably, the term of office R in step S2 is calculated as follows:

[0020]

[0021] Once a CH (Chancellor) is determined, the cluster does not select a new CH during its term. A new CH is selected when the term ends or when the cluster head energy is exhausted during the term. The term is determined by the remaining energy. Within the CH selection region, a node is elected as the CH when the remaining energy is greater than a threshold; when the remaining energy is less than the threshold, the node enters a dormant state. When the CH term ends, regardless of whether the remaining energy is higher than E (E), the remaining energy remains constant. th,CH CH will enter hibernation mode.

[0022] Preferably, in step S3,

[0023] During the listening phase, the CM node is responsible for spectrum sensing, environmental monitoring, and energy harvesting. If it detects idle spectrum, the node also transmits data. During the sleep phase, the node is only responsible for energy harvesting. When the node's energy is greater than E... th,2 A node only starts working when its remaining energy is greater than the energy required to complete one full data cycle. The node's threshold depends on whether its remaining energy is greater than the energy required to complete one full data cycle. The node thresholds are categorized as follows:

[0024]

[0025] Based on two thresholds, CM nodes are divided into three working modes:

[0026]

[0027] During normal operation at the working period node, the energy consumption is E. CM,i Energy consumption during the listening period is:

[0028] E CM,i =N Ss E Ss +N Em E Em +N Em E Df ;

[0029] During the dormant period, the node is assumed to be in an ideal state with zero energy consumption.

[0030] E CM,i =0;;

[0031] During the operational and monitoring periods, if a node detects abnormal data, it immediately marks it as an abnormal data packet and uploads the monitored data.

[0032] The number of times a node detects data during its operational period is determined by its remaining energy. Besides detecting the surrounding environment, the most important task of a node is to upload the detection results to terminal equipment for data analysis, enabling remote monitoring of oil and gas pipelines. The maximum number of environmental detections N for a node is [not specified]. Em The calculation is as follows:

[0033]

[0034] E work This represents the energy consumption required for a node to function normally.

[0035] E work =E Ss +E Em +E Dt +E Df ;

[0036] During the working and listening periods, the number of times a node can listen to the spectrum is limited. A node must ensure that its remaining energy is sufficient to upload data; otherwise, the listening becomes meaningless. The maximum number of times a node can listen to the spectrum is N. Ss The calculation is as follows:

[0037]

[0038] Preferably, in step S4,

[0039] The formula for expressing the degree of deviation of data from the mean is as follows:

[0040]

[0041] Grde i,j This represents the degree of deviation of data in the j-th node of the i-th cluster. Using the 3σ principle, data with a deviation exceeding 3σ from the standard monitoring data is considered abnormal. i,j A larger value indicates more important data, as it suggests that there may have been anomalies in the detection environment.

[0042] Abnormal data is represented as follows:

[0043]

[0044] The specific operation of step S4 is as follows:

[0045] (5) Within one time cycle, the sensor acquires and stores the acquired data.

[0046] (6) The sensor processes the monitored data and calculates Grde. i,j ,

[0047] (7) Determine Grde i,j If the value is greater than 3σ, upload the outlier immediately; otherwise, upload the latest detection data.

[0048] (8) Abnormal data is marked as abnormal before being uploaded, and CH also prioritizes relaying abnormal data.

[0049] Preferably, in step S5, the filtering and fusion of data between clusters is specifically divided into:

[0050] Data fusion is divided into intra-cluster fusion and inter-cluster fusion. In an oil and gas IoT monitoring environment, sensors collect various physical quantities, such as temperature, humidity, and pressure. The data from sensor nodes is represented by X... i,n ={X i,1 ,X i,2 ,....,X i,n The importance level of sensor data is represented by Y. i,n={Y i,1 ,Y i,2 ,....,Y i,n} represents the data fusion scheme, where i represents the sensor type and n represents the number of data points. The data fusion scheme includes two aspects: intra-cluster data fusion and inter-cluster data fusion.

[0051] A. Intra-cluster data fusion

[0052] Cluster member nodes will collect monitoring data X i,n Upload to the cluster head and label the Y value of each data point. i,n After completing data collection, the cluster head node classifies the data and selects the cluster with the largest Y value in each category. i,n The cluster head will use the Y-axis of this data set. n Compare and select the largest Y. n As an important level of data packets,

[0053] TY i,n =max(Y i,1 ,Y i,2 ,....,Y i,n );

[0054] TY i,n →X i,n ={X i,1 ,X i,2 ,....,X i,n};

[0055] PA = max(TY1, TY2, ..., TY) i );

[0056] B. Inter-cluster data fusion

[0057] After completing intra-cluster data fusion, the relay cluster head compares the fused intra-cluster data with the fused inter-cluster data and selects the data packet with the highest PA (Patent Amount) for uploading. This ensures that each cluster head only uploads necessary data during operation, automatically filtering redundant or useless data. To prevent the loss of important data, an importance level threshold is set. If the PA is greater than the threshold, fusion is unnecessary and the data is uploaded directly; if all PAs are greater than the threshold, all data is uploaded. The importance level threshold is set as follows: Data... th As shown below,

[0058]

[0059]

[0060] ② Maximum number of transmissions per node: In the oil and gas IoT, nodes perform different functions, resulting in different energy consumption. Therefore, energy consumption patterns of network nodes are modeled and analyzed. The network has two different energy consumption patterns: CH and CM. It is assumed that all nodes have the capabilities to monitor the surrounding environment, wirelessly collect energy, collaboratively sense the spectrum, receive / transmit data, and filter data. CMs are responsible for monitoring pipeline status, sensing spectrum information, collecting energy, filtering the collected information, and transmitting monitoring data to CHs. CHs are responsible for sensing spectrum information and receiving / relaying data.

[0061] The energy consumption of a CM mainly consists of spectrum sensing, environmental monitoring, data transmission, and data filtering, with data transmission accounting for almost all of the energy consumption.

[0062] E CM,i =N Ss E Ss +N Em E Em +E Dt +N Em E Df ;

[0063] In the formula E CM,i E represents the energy consumed by a node. Ss E represents the energy consumption for spectrum sensing. Em E represents the energy of the monitored environment. Dt E represents the energy of uploading data. Df N represents the energy required for data filtering. Em N represents the number of times the environment was monitored, and the monitored data was filtered. Ss Represents the number of spectrum sensing operations, where E Dt =E (k,d) ,

[0064] The energy consumed by the CH node is:

[0065] E CH,i =E Ss +E Dr +E Dt +E Dp ;

[0066] Where E CH,i E represents the energy consumed by a node. Ss E represents the energy consumption for spectrum sensing. Dr E represents the energy consumed in receiving data. Dt E represents the energy consumed in uploading data. Dp This indicates the energy consumed in data processing.

[0067] Compared with existing technologies, the beneficial effects of this invention are: this oil and gas IoT energy management method based on data fusion can reduce the communication distance between sensor nodes and the aggregation node, thereby improving network energy efficiency. Furthermore, it can reduce or avoid the energy consumption of redundant information. The algorithm can determine the degree of data anomaly, avoiding duplication of network information and excessive energy consumption. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the data fusion mechanism of the present invention;

[0069] Figure 2 This is a schematic diagram of the remaining energy of the network in this invention;

[0070] Figure 3 This is a schematic diagram showing the remaining energy of a network node at 5000 frames. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] Please see Figure 1-3 This invention provides a technical solution: an oil and gas Internet of Things energy management method based on data fusion.

[0073] This efficient energy consumption management algorithm for oil and gas IoT includes the following steps:

[0074] Step 1: Selecting the Cluster Head (CH) for the Network Cluster:

[0075] Step 101: Divide the core region and the edge region;

[0076] Step 102: All nodes in the network calculate and store their remaining energy;

[0077] Step 103: Select the node with the most remaining energy in the core region as the cluster head;

[0078] Once the cluster head is selected, its term of office is calculated.

[0079] ① Select nodes that meet the remaining energy criterion. The calculation formula is described below:

[0080] E th,res >ω×E CH,i ,(ω=3);

[0081] E th,CH =Eth,res ;

[0082] Among them, E th,res E represents the remaining energy of CH. CH,i E represents the energy consumed by a CH node during one normal operation. th,CH This represents the energy threshold of CH.

[0083] ② Once the CH (Cluster Head) is determined, a term R is set. During the term, the cluster does not select a new CH. When the term ends or the cluster head energy is exhausted during the term, a new CH is selected. The term is determined by the remaining energy. The term R is calculated as follows:

[0084]

[0085] This document stipulates that within the CH selection region, a node is eligible to be elected as the CH only when the remaining energy is greater than a threshold; when the remaining energy is less than the threshold, the node is in a dormant state. When the CH term ends, regardless of whether the remaining energy is higher than E... th,CH CH will enter hibernation mode.

[0086] Step 2: The node begins transmitting and forwarding data, and the cluster head begins receiving data.

[0087] Step 201: The node sends RTS information to the cluster head;

[0088] Step 202: The cluster head sends CTS information to the node that sent the RTS message;

[0089] Step 203: The cluster head sends an ACK message to the node that received the RTS message to confirm its identity.

[0090] Once confirmation is complete, the node begins transmitting data.

[0091] ① The node filters the data type;

[0092] Since the Z-score data distribution follows a "normal distribution," it is a method used for data normalization. The calculation formula is as follows:

[0093]

[0094] To facilitate the screening of monitoring data, the formula was improved to represent the degree of deviation of the data from the mean.

[0095]

[0096] Grde i,j This represents the degree of deviation of the data in the j-th node of the i-th cluster. Using the 3σ principle, data deviating more than 3σ from the standard monitoring data is considered abnormal. i,jA larger value indicates more important data, as it suggests potential anomalies in the detection environment. Abnormal data is represented as follows:

[0097]

[0098] (1) Within a time cycle, the sensor acquires data and stores the acquired data.

[0099] (2) The sensor processes the monitored data and calculates Grde. i,j .

[0100] (3) Determine Grde i,j If the value is greater than 3σ, upload the outlier immediately; otherwise, upload the latest detection data.

[0101] (4) Abnormal data is marked as abnormal before being uploaded, and CH also prioritizes relaying abnormal data.

[0102] like Figure 1 As shown, data fusion is divided into intra-cluster fusion and inter-cluster fusion, as illustrated in the diagram below. Figure 1 In oil and gas IoT monitoring environments, sensors collect various physical quantities, such as temperature, humidity, and pressure. The data from the sensor nodes is represented by X... i,n ={X i,1 ,X i,2 ,....,X i,n The importance level of sensor data is represented by Y. i,n ={Y i,1 ,Y i,2 ,....,Y i,n Let} represent the data fusion scheme, where i represents the sensor type and n represents the number of data points. The data fusion scheme includes two aspects: intra-cluster data fusion and inter-cluster data fusion.

[0103] A. Intra-cluster data fusion

[0104] Cluster member nodes will collect monitoring data X i,n Upload to the cluster head and label the Y value of each data point. i,n After completing data collection, the cluster head node classifies the data and selects the cluster with the largest Y value in each category. i,n The cluster head will store the Y-axis data of this set of data. n Compare and select the largest Y. n As an important level of data packets.

[0105] TY i,n =max(Y i,1 ,Y i,2 ,....,Y i,n );

[0106] TYi,n →X i,n ={X i,1 ,X i,2 ,....,X i,n};

[0107] PA = max(TY1, TY2, ..., TY) i );

[0108] B. Inter-cluster data fusion

[0109] After completing intra-cluster data fusion, the relay cluster head compares the fused intra-cluster data with the fused inter-cluster data and selects the data packet with the highest PA (Patent Amount) for uploading. This ensures that each cluster head only uploads necessary data during operation, automatically filtering out redundant or useless data. To prevent the loss of important data, an importance level threshold is set. If the PA is greater than the threshold, fusion is unnecessary and the data is uploaded directly; if all PAs are greater than the threshold, all data is uploaded. The importance level threshold is set for Data. th As shown below.

[0110]

[0111]

[0112] ② Maximum number of node transmissions. In the oil and gas IoT, nodes perform different functions, resulting in different energy consumption. Therefore, energy consumption patterns of network nodes are modeled and analyzed. The network has two different energy consumption modes for nodes: CH and CM. It is assumed that all nodes have the capabilities to monitor the surrounding environment, wirelessly harvest energy, collaboratively sense the spectrum, receive / transmit data, and filter data. CMs are responsible for monitoring pipeline status, sensing spectrum information, harvesting energy, filtering the harvested information, and transmitting monitoring data to CHs. CHs are responsible for sensing spectrum information and receiving / relaying data.

[0113] The energy consumption of a CM mainly consists of spectrum sensing, environmental monitoring, data transmission, and data filtering, with data transmission accounting for almost all of the energy consumption.

[0114] E CM,i =N Ss E Ss +N Em E Em +E Dt +N Em E Df ;

[0115] In the formula E CM,i E represents the energy consumed by a node. Ss E represents the energy consumption for spectrum sensing. Em E represents the energy of the monitored environment.Dt E represents the energy of uploading data. Df N represents the energy required for data filtering. Em This indicates the number of times the environment was monitored, and the monitored data was filtered. N Ss Represents the number of spectrum sensing operations, where E Dt =E (k,d)

[0116] The energy consumed by the CH node is:

[0117] E CH,i =E Ss +E Dr +E Dt +E Dp ;

[0118] Where E CH,i E represents the energy consumed by a node. Ss E represents the energy consumption for spectrum sensing. Dr E represents the energy consumed in receiving data. Dt E represents the energy consumed in uploading data. Dp This indicates the energy consumed in data processing.

[0119] ③ To conserve node energy and prevent insufficient energy during data transmission, the node's operating modes are defined as working period, listening period, and sleep period. During the working period, the node operates normally, performing tasks such as energy harvesting, environmental monitoring, spectrum sensing, and data uploading. During the listening period, the CM node is responsible for spectrum sensing, environmental monitoring, and energy harvesting; if idle spectrum is detected, the node also transmits data. During the sleep period, the node is only responsible for energy harvesting. When the node's energy is greater than E... th,2 A node only begins working when its remaining energy is greater than the energy required to complete one full data cycle. Node thresholds are categorized as follows:

[0120]

[0121] Based on two thresholds, CM nodes are divided into three working modes:

[0122]

[0123] During normal operation at the working period node, the energy consumption is E. CM,i Energy consumption during the listening period is:

[0124] E CM,i =N Ss E Ss +N Em E Em +NEm E Df ;

[0125] During the dormant period, the node is assumed to be in an ideal state with zero energy consumption.

[0126] E CM,i =0;

[0127] During the operational and monitoring periods, if a node detects abnormal data, it immediately marks it as an abnormal data packet and uploads the monitored data.

[0128] The number of times a node performs environmental data checks during its operational period is determined by its remaining energy. Besides monitoring its surroundings, the most important task of a node is to upload the results to terminal equipment for data analysis, enabling remote monitoring of oil and gas pipelines. The maximum number of environmental checks N performed by a node is [not specified in the original text]. Em The calculation is as follows:

[0129]

[0130] E work This represents the energy consumption required for a node to function normally.

[0131] E work =E Ss +E Em +E Dt +E Df ;

[0132] During operation and listening periods, the number of times a node can listen to the spectrum is limited. A node must ensure that its remaining energy is sufficient to upload data; otherwise, listening becomes meaningless. The maximum number of times a node can listen to the spectrum is N. Ss The calculation is as follows:

[0133]

[0134] Examples of confidentiality experiments:

[0135] MATLAB 2018a was used as the simulation platform to evaluate the network performance using the proposed protocol. The network was built within a 200m long and 100m wide area, with relay nodes randomly distributed within a 100×200m range. Each node was located within a 1×1m grid, ensuring that the coordinates of each node were known. A sink node was placed on the right side of the area, with coordinates (200, 200). To provide sufficient energy to the nodes, two energy emitters were used, with coordinates (0, 200) and (200, 0), respectively.

[0136] To demonstrate the simulation performance of our proposed scheme, we compare it with the EEM-EHWSN and EB-LEACH schemes.

[0137] 1. Comparison of remaining energy in network nodes

[0138] Figure 2 As shown, at 5000 frames, in the EB-LEACH scheme, the energy of one node has been completely exhausted, and the energy level is maintained at 2*10. 10 ~3*10 10 The EEM-EHWSN and DSF-EMA schemes, due to the alternation of nodes between working and sleeping periods, ensure timely energy supply and show little change compared to the 1000 frames per second scheme.

[0139] In summary, the energy management method proposed in this invention can effectively increase network lifetime, and its performance is superior to the previously proposed EEM-EHWSN and EB-LEACH schemes.

[0140] The above description is merely a preferred embodiment of the present invention, but the present invention should not be limited to the content disclosed in this embodiment and the accompanying drawings. Therefore, any equivalent or modified embodiments made without departing from the technical solutions disclosed in the present invention fall within the protection scope of the present invention.

[0141] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for energy management in the oil and gas Internet of Things based on data fusion, characterized in that, Includes the following steps: S1. Divide the core region and the edge region based on the size of the cluster. Select the cluster head in the core region by calculation. In this process, all nodes in the network need to calculate and store their remaining energy. Select the node with the most remaining energy in the core region as the cluster head. S2. After selecting the cluster head, calculate the cluster head term based on the remaining energy. S3. Nodes in the cluster that are not selected as cluster heads calculate their working modes based on their remaining energy and classify them into working modes: working period, listening period, and dormant period. During the working period, the node is in normal working condition, completing tasks such as energy harvesting, environmental monitoring, spectrum sensing, and data uploading; During the listening period, the node is responsible for spectrum sensing, environmental monitoring, and energy harvesting. If it senses that the spectrum is idle, the node will also transmit data. During the dormant period, the node is only responsible for energy collection; During the listening phase, the CM node is responsible for spectrum sensing, environmental monitoring, and energy harvesting. If it detects idle spectrum, the node also transmits data. During the dormant phase, the node is only responsible for energy harvesting. When the node's energy is greater than... A node only starts working when its remaining energy is greater than the energy required to complete one full data cycle. The node's threshold depends on whether its remaining energy is greater than the energy required to complete one full data cycle. The node thresholds are categorized as follows: ; Based on two thresholds, CM nodes are divided into three working modes: ; During normal operation at key work periods, energy consumption is... Energy consumption during the listening period is: ; During the dormant period, the node is assumed to be in an ideal state with zero energy consumption. ; During the operational and monitoring periods, if a node detects abnormal data, it immediately marks it as an abnormal data packet and uploads the monitored data. The number of times a node performs data checks during its operational period is determined by its remaining energy. Besides monitoring its surroundings, the node's most important task is to upload the results to terminal equipment for data analysis, enabling remote monitoring of oil and gas pipelines. The maximum number of environmental checks a node can perform is [not specified in the original text]. The calculation is as follows: ; This represents the energy consumption required for a node to function normally. ; During operation and listening periods, the number of times a node can listen to the spectrum is limited. A node must ensure that its remaining energy is sufficient to upload data; otherwise, listening becomes meaningless. The maximum number of times a node can listen to the spectrum is limited. The calculation is as follows: ; S4. Data in the cluster that is not selected as the cluster head is filtered according to the data filtering principle. Only important data is selected for uploading. The data filtering is divided according to the degree of deviation of the data. The greater the degree of deviation, the more abnormal the detection environment is. S5. Data filtering and fusion between clusters; Data fusion is divided into intra-cluster fusion and inter-cluster fusion. In the oil and gas IoT monitoring environment, sensors collect different physical quantities, including temperature, humidity, and pressure; the data from sensor nodes are used... The importance level of sensor data is indicated by... This indicates that i represents the sensor type and n represents the number of data points. The data fusion scheme includes two aspects: intra-cluster data fusion and inter-cluster data fusion. A. Intra-cluster data fusion: Cluster member nodes will collect monitoring data Upload to the cluster head and label each data item. After completing data collection, the cluster head node classifies the data and selects the cluster with the largest value in each category. The cluster head will store this set of data. Compare and select the largest. As an important level of data packets, ; ; ; B. Inter-cluster data fusion: After completing the intra-cluster data fusion, the relay cluster head compares the intra-cluster fused data with the inter-cluster fused data and selects the data packet with the highest PA for uploading. In this way, each cluster head only uploads the necessary data during operation and automatically filters redundant or useless data. In order to avoid the loss of important data, an importance level threshold is set. If the PA is greater than the threshold, no fusion is required and the data is uploaded directly. If all data are greater than the threshold, all data are uploaded. The maximum number of transmissions per node varies depending on the function performed by the node in the oil and gas Internet of Things (IoT). Therefore, energy consumption patterns of network nodes are modeled and analyzed. The network has two different energy consumption modes: CH (Chain Controller) and CM (Community Controller). All nodes are assumed to have the capabilities to monitor the surrounding environment, wirelessly harvest energy, collaboratively sense the spectrum, receive / transmit data, and filter data. The CM is responsible for monitoring pipeline status, sensing spectrum information, harvesting energy, filtering the harvested information, and transmitting monitoring data to the CH. The CH is responsible for sensing spectrum information and receiving / relaying data. The energy consumption of the CM node is calculated as follows: ; In the formula This represents the energy consumed by the node. This represents the energy consumption for spectrum sensing. This indicates the energy level of the monitored environment. This represents the energy of the uploaded data. This represents the energy required for data filtering. This indicates the number of times the environment was monitored, and the monitored data is filtered. This indicates the number of times the spectrum was sensed, where, ; The energy consumed by the CH node is: ; in This represents the energy consumed by the node. This represents the energy consumption for spectrum sensing. This indicates the energy consumed in receiving data. This indicates the energy consumed during data upload. This indicates the energy consumed in data processing.

2. The oil and gas Internet of Things energy management method based on data fusion according to claim 1, characterized in that: The specific formula for calculating the remaining energy of the node in step S1 is as follows: ; ; Where CH represents the cluster head, This represents the remaining energy of CH. This represents the energy consumed by a CH node during one normal operation. This represents the energy threshold of CH.

3. The oil and gas Internet of Things energy management method based on data fusion according to claim 1, characterized in that: The term of office R in step S2 is calculated as follows: ; Once a CH (Chancellor) is determined, the cluster does not select a new CH during its term. A new CH is selected when the term ends or when the cluster head's energy is depleted during the term. The term is determined by the remaining energy. Within the CH selection region, a node is elected as CH when the remaining energy is greater than a threshold; when the remaining energy is less than the threshold, the node enters a dormant state. When the CH term ends, regardless of whether the remaining energy is higher than or lower than the threshold... CH will enter hibernation mode.

4. The oil and gas Internet of Things energy management method based on data fusion according to claim 1, characterized in that: In step S4, the formula for the degree of deviation of the data from the mean is as follows: ; This indicates the degree of deviation of the data in the j-th node of the i-th cluster, using... In principle, if the deviation between the monitoring data and the standard monitoring data exceeds [a certain threshold], [the following conditions apply]. This is abnormal data. The larger the value, the more important the data, because it indicates that there may be an anomaly in the detection environment; Step S4 is performed as follows: Within one time loop, the sensor acquires and stores the acquired data. The sensor processes the monitored data and calculates... , judge Is it greater than If so, upload the abnormal value immediately; Otherwise, upload the latest test data. Abnormal data is marked as abnormal before being uploaded, and CH also prioritizes relaying abnormal data.