Green energy fault monitoring method and system based on computing network

Abstract

The invention relates to the technical field of green energy power generation, in particular to a green energy fault monitoring method and system based on an arithmetic network, and the system comprises a data collection layer, an arithmetic network resource scheduling layer, an intelligent analysis layer, and a fault response layer. The algorithm resource scheduling layer adjusts the calculation logic of a calculation network according to the output power of the green energy power generation network, the data acquisition layer transmits collected data to the intelligent analysis layer, the intelligent analysis layer outputs an analysis result to the fault response layer, and the fault response layer communicates with the outside. According to the invention, the calculated threshold value of each node is compared with the real-time data of each node through the fault response layer, and a four-level alarm system is formed, so that classification of different fault conditions is realized, and comprehensive prevention of small faults without shutdown and large faults is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of green energy power generation, and in particular to a green energy fault monitoring method and system based on a computing network. Background Art

[0002] Green energy refers to generator sets that use renewable energy sources such as wind, solar, and hydropower to generate electricity. These generator sets usually use clean energy, do not produce air pollution and greenhouse gas emissions, and are environmentally friendly. Green energy can help reduce dependence on fossil fuels and promote sustainable development. Common green energy sources include wind turbines, solar turbines, and hydropower generators. With the improvement of environmental awareness and technological advancements, green energy has been widely used and developed around the world.

[0003] In order to ensure that the green energy power generation grid is in an effective working state, it is necessary to monitor the operating status of the green energy power generation grid. However, the existing monitoring system does not monitor faults in a timely manner, and when a fault occurs, it needs to be shut down for maintenance, which affects the power generation efficiency. Summary of the Invention

[0004] In order to solve the problem that the existing monitoring system cannot monitor faults in a timely manner and needs to be shut down for maintenance when a fault occurs, the present invention proposes a green energy fault monitoring system based on a computing network.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A green energy fault monitoring system based on a computing network includes a data acquisition layer, a computing network resource scheduling layer, an intelligent analysis layer, and a fault response layer. The data acquisition layer includes multiple sensors deployed at each energy node. The algorithm resource scheduling layer adjusts the computing logic of the computing network according to the output power of the green energy power generation network. The data acquisition layer transmits the collected data to the intelligent analysis layer. The intelligent analysis layer outputs the analysis results to the fault response layer, and the fault response layer communicates with the outside.

[0006] Furthermore, the green energy power generation grid includes solar power generation units, wind power generation units and hydropower generation units. The solar power generation unit includes output voltage sensor 1, output current sensor 1, solar panel temperature sensor and light intensity sensor. The wind power generation unit includes output voltage sensor 2, output current sensor 2 and wind turbine vibration acceleration sensor. The hydropower generation unit includes output voltage sensor 3, output current sensor 3 and generator vibration acceleration sensor.

[0007] A green energy fault monitoring method based on computing network collects the operating data of each part of the green energy power generation network through the data collection layer, cleans, standardizes and aggregates the collected data, and selects the system operation mode according to the output power of the green energy power generation network. The operation mode is divided into high green energy mode and low green energy mode. In high green energy mode, tasks are processed according to the size of the dynamic weight of the task. In low green energy mode, except for necessary tasks, all other tasks are migrated to other energy nodes; the threshold value of each node is calculated by the data collected by the sensor through the intelligent analysis layer. The calculation formula for the threshold of each node is: W1=μ1±σ0; σ0=μ1(1-σ1); ; The fault response layer then compares the calculated thresholds of each node with the real-time data of each node, and forms a four-level alarm system, which includes early warning mode, minor alarm mode, severe alarm mode, and disaster mode. When μ1-σ0<W0<μ1+σ0, the system is judged to be in warning level mode, and the operation and maintenance work order is triggered in the warning level mode; When μ1-2σ0<W0<μ1-σ0 or μ1+σ0<W0<μ1+2σ0, the system is judged to be in mild alarm mode. In mild alarm mode, the power output is limited to 80% and the backup heat dissipation is activated; When μ1-3σ0<W0<μ1-2σ0 or μ1+2σ0<W0<μ1+3σ0, the system is judged to be in severe alarm mode, in which emergency shutdown is carried out and fault location information is pushed; When μ1-4σ0<W0<μ1-3σ0 or μ1+3σ0<W0<μ1+4σ0, the system is judged to be in disaster-level mode. In disaster-level mode, the DC side circuit breaker is cut off and the fire protection system is started.

[0008] Furthermore, necessary tasks are analytical calculations related to fault monitoring.

[0009] Furthermore, the normalized data is calculated using the following formula: ; Data aggregation is performed by taking a sliding average of data at multiple time points; In the data cleaning step, the 3-times standard deviation method is used to remove outliers. A data point x is considered an outlier if it meets the following conditions: |x−μ0|>3σ2; The sliding average processing is calculated by the sliding average calculation formula: .

[0010] Furthermore, the judgment logic of the operation mode is: when P 总 >P 阈值 , the system is in high green energy mode; when P 总 ≤P 阈值 , the system is in low green energy mode.

[0011] Furthermore, other energy nodes are thermal power generation nodes.

[0012] Furthermore, the task dynamic weight calculation formula is: W i =α⋅S i +β⋅(1−U cpu )+γ⋅B n .

[0013] Furthermore, the network resource scheduling strategy of the network resource scheduling layer is: ① preemptive scheduling: high-weight green energy nodes can preempt tasks of low-weight grid nodes; ② elastic resource reservation: reserve some green energy node resources for sudden fault monitoring tasks.

[0014] Furthermore, some green energy node resources account for 25% of the overall green energy node resources.

[0015] Beneficial effects: The present invention compares the calculated threshold of each node with the real-time data of each node through the fault response layer, and forms a four-level alarm system, which realizes the classification of different fault situations and achieves comprehensive prevention of minor faults without stopping the machine and major faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the top structure of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0018] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0019] Reference Figure 1A green energy fault monitoring system based on a computing network includes a data acquisition layer, a computing network resource scheduling layer, an intelligent analysis layer, and a fault response layer. The data acquisition layer includes a variety of sensors deployed at each energy node. The algorithm resource scheduling layer adjusts the computing logic of the computing network according to the output power of the green energy power generation network. The data acquisition layer transmits the collected data to the intelligent analysis layer. The intelligent analysis layer outputs the analysis results to the fault response layer, and the fault response layer communicates with the outside.

[0020] The data collection layer includes multi-source sensors. Several multi-source sensors are deployed in the green energy power generation grid. These sensors include output voltage sensor 1, output current sensor 1, solar panel temperature sensor, light intensity sensor, output voltage sensor 2, output current sensor 2, wind turbine vibration acceleration sensor, output voltage sensor 3, output current sensor 3, and generator vibration acceleration sensor. These sensors collect voltage, current, temperature, light intensity, wind speed, and water flow velocity data from the power generation grid, and divide the collected data into private and public datasets. The green energy power generation grid includes solar power generation units, wind power generation units, and hydropower generation units. Solar power generation units include output voltage sensor 1, output current sensor 1, solar panel temperature sensor, and light intensity sensor. Wind power generation units include output voltage sensor 2, output current sensor 2, and wind turbine vibration acceleration sensor. Hydropower generation units include output voltage sensor 3, output current sensor 3, and generator vibration acceleration sensor.

[0021] A green energy fault monitoring method based on a computing network collects operating data from various parts of a green energy power generation network through a data acquisition layer, and cleans, standardizes, and aggregates the collected data, wherein data cleaning is accomplished by removing outliers or missing values; Normalized data are calculated using the following formula: ; Where x is the original data, μ0 is the mean, and σ2 is the standard deviation; Data aggregation is performed by taking a sliding average of data at multiple time points; In the data cleaning step, the 3-times standard deviation method is used to remove outliers. A data point x is considered an outlier if it meets the following conditions: |x−μ0|>3σ2; The sliding average processing is calculated by the sliding average calculation formula: .

[0022] Among them, MA t is the sliding average of time t, n is the sliding window size, X t-i is the sampling data of the previous n time points.

[0023] The system's operating mode is selected based on the output power of the green energy power generation grid. The operating modes are divided into high green energy mode and low green energy mode. In high green energy mode, tasks are processed based on the size of the dynamic weight of the tasks. In low green energy mode, except for necessary tasks, all other tasks are migrated to other energy nodes.

[0024] Necessary tasks are analytical calculations related to fault monitoring.

[0025] The judgment logic of the operation mode is: when P 总 >P 阈值 , the system is in high green energy mode; when P 总 ≤P 阈值 , the system is in low green energy mode.

[0026] P 总 is the output power of the green energy power grid, P 阈值 is the threshold power of the green energy power grid.

[0027] Other energy nodes are thermal power generation nodes.

[0028] The calculation formula of the task dynamic weight is: W i =α⋅S i +β⋅(1−U cpu )+γ⋅B n .

[0029] S i is the proportion of green energy (0~1); U cpu is the CPU utilization; B n is the available bandwidth of the node; dynamic weight W i High priority tasks are assigned and processed.

[0030] The network resource scheduling strategies of the network resource scheduling layer are as follows: ① preemptive scheduling: high-weight green energy nodes can preempt tasks of low-weight grid nodes; ② elastic resource reservation: reserve some green energy node resources for sudden fault monitoring tasks.

[0031] Partial green energy node resources account for 25% of the total green energy node resources.

[0032] The intelligent analysis layer calculates the threshold of each node based on the data collected by the sensor. The calculation formula for the threshold of each node is: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0033] W1: threshold; σ0: fluctuation error; μ1: mean of historical data; σ1: error volatility; ε t−1 : The residual at the previous moment (such as photovoltaic current deviation, abnormal value of wind turbine vibration acceleration); α0: long-term average volatility (determined by historical equipment data statistics); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging).

[0034] σ t−1 : Error volatility at the previous moment.

[0035] The calculated thresholds of each node are compared with the real-time data of each node to form a four-level alarm system, which includes early warning mode, mild alarm mode, severe alarm mode, and disaster mode. When μ1-σ0<W0<μ1+σ0, the system is judged to be in warning level mode, and the operation and maintenance work order is triggered in the warning level mode; When μ1-2σ0<W0<μ1-σ0 or μ1+σ0<W0<μ1+2σ0, the system is judged to be in mild alarm mode. In mild alarm mode, the power output is limited to 80% and the backup heat dissipation is activated; When μ1-3σ0<W0<μ1-2σ0 or μ1+2σ0<W0<μ1+3σ0, the system is judged to be in severe alarm mode, in which emergency shutdown is carried out and fault location information is pushed; When μ1-4σ0<W0<μ1-3σ0 or μ1+3σ0<W0<μ1+4σ0, the system is judged to be in disaster-level mode. In disaster-level mode, the DC side circuit breaker is cut off and the fire protection system is started.

[0036] The solar power generation unit includes the real-time detection value A1 of the output voltage sensor 1, the real-time detection value A2 of the output current sensor 1, the real-time detection value A3 of the solar panel temperature sensor, and the real-time detection value A4 of the light intensity sensor; the wind power generation unit includes the real-time detection value B1 of the output voltage sensor 2, the real-time detection value B2 of the output current sensor 2, and the real-time detection value B3 of the wind turbine vibration acceleration sensor; the hydropower generation unit includes the real-time detection value C1 of the output voltage sensor 3, the real-time detection value C2 of the output current sensor 3, and the real-time detection value C3 of the generator vibration acceleration sensor.

[0037] ① When calculating the output voltage threshold of the solar power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0038] W1: The threshold value of the detection data of the output voltage sensor 1; σ0: Fluctuation error of the real-time voltage detection value A1 of the output voltage sensor 1; μ1: the mean value of the historical data of the real-time detection value A1 of the output voltage sensor 1; σ1: Error fluctuation rate of the detection data of output voltage sensor 1; ε t−1 : Residual error at the previous moment (voltage of output voltage sensor 1); α0: long-term average volatility (determined by historical equipment data of solar power generation units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of solar power generation units).

[0039] ② When calculating the output current threshold of the solar power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0040] W1: The threshold value of the detection data of the output current sensor 1; σ0: Fluctuation error of the real-time detection value A2 of the output current sensor 1; μ1: the average value of the historical data of the real-time detection value A2 of the output current sensor 1; σ1: The error fluctuation rate of the current of output current sensor 1; ε t−1 : The residual error at the previous moment (the current of output current sensor 1); α0: long-term average volatility (determined by historical equipment data of solar power generation units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of solar power generation units).

[0041] ③ When calculating the threshold value of the solar panel temperature of the solar power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0042] W1: The threshold of the detection data of the solar panel temperature sensor; σ0: Fluctuation error of the real-time temperature detection value A3 of the solar panel temperature sensor; μ1: the mean of the historical data of the real-time temperature detection value A3 of the solar panel temperature sensor; σ1: Error fluctuation rate of detection data of solar panel temperature sensor; ε t−1 : The residual of the previous moment (the temperature of the solar panel temperature sensor); α0: long-term average volatility (determined by historical equipment data of solar power generation units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of solar power generation units).

[0043] ④ When calculating the threshold value of light intensity of solar power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0044] W1: The threshold of the detection data of the light intensity sensor; σ0: Fluctuation error of the real-time light intensity detection value A4 of the light intensity sensor; μ1: the mean of the historical data of the light intensity sensor’s real-time detection value A4; σ1: Error fluctuation rate of the detection data of the light intensity sensor; ε t−1 : The residual error at the previous moment (light intensity of the light intensity sensor); α0: long-term average volatility (determined by historical equipment data of solar power generation units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of solar power generation units).

[0045] ⑤ When calculating the output voltage threshold of the wind power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0046] W1: The threshold value of the detection data of the output voltage sensor 2; σ0: Fluctuation error of the real-time voltage detection value B1 of the output voltage sensor 2; μ1: the mean value of the historical data of the real-time voltage detection value B1 of the output voltage sensor 2; σ1: Error fluctuation rate of the detection data of output voltage sensor 2; ε t−1 : The residual error between the first two moments (the voltage of the second output voltage sensor); α0: long-term average volatility (determined by historical equipment data statistics of wind power units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of wind power units).

[0047] ⑥ When calculating the output current threshold of the wind power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0048] W1: The threshold value of the detection data of the output current sensor 2; σ0: Fluctuation error of the real-time detection value B2 of the output current sensor 2; μ1: the average value of the historical data of the real-time detection value B2 of the output current sensor 2; σ1: The error fluctuation rate of the current of the output current sensor 2; ε t−1 : The residual error between the first two moments (the current of output current sensor 2); α0: long-term average volatility (determined by historical equipment data statistics of wind power units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of wind power units).

[0049] ⑦ When calculating the threshold value of the wind turbine vibration acceleration of the wind power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0050] W1: Threshold of detection data of fan vibration acceleration sensor; σ0: Fluctuation error of the fan vibration acceleration real-time detection value B3 of the fan vibration acceleration sensor; μ1: the mean value of the historical data of the fan vibration acceleration real-time detection value B3 of the fan vibration acceleration sensor; σ1: Error fluctuation rate of the detection data of the fan vibration acceleration sensor; ε t−1 : The residual between the first two moments (the fan vibration acceleration of the fan vibration acceleration sensor); α0: long-term average volatility (determined by historical equipment data statistics of wind power units); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging of wind power units).

[0051] ⑧ When calculating the threshold value of the output voltage of the hydroelectric power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0052] W1: The threshold of the detection data of the output voltage sensor 3; σ0: Fluctuation error of the real-time voltage detection value C1 of output voltage sensor 3; μ1: the mean value of the historical data of the real-time voltage detection value C1 of the output voltage sensor 3; σ1: error fluctuation rate of the detection data of output voltage sensor 3; ε t−1 : The residual of the first three moments (the voltage of output voltage sensor 3); α0: long-term average volatility (determined by historical equipment data statistics of the hydropower unit); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging in hydropower units).

[0053] ⑨ When calculating the output current threshold of the hydroelectric power generation unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0054] W1: The threshold value of the detection data of the output current sensor 3; σ0: Fluctuation error of the real-time detection value C2 of the output current sensor 3; μ1: the average value of the historical data of the real-time detection value C2 of the output current sensor 3; σ1: the error fluctuation rate of the current of the output current sensor 3; ε t−1 : The residual of the first three moments (the current of output current sensor 3); α0: long-term average volatility (determined by historical equipment data statistics of the hydropower unit); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging in hydropower units).

[0055] ⑩ When calculating the threshold value of the fan vibration acceleration of the hydropower unit: W1=μ1±σ0; σ0=μ1(1-σ1); .

[0056] W1: Threshold of detection data of generator vibration acceleration sensor; σ0: Fluctuation error of the real-time detection value C3 of the generator vibration acceleration sensor; μ1: the mean value of the historical data of the real-time detection value C3 of the generator vibration acceleration sensor; σ1: Error fluctuation rate of detection data of generator vibration acceleration sensor; ε t−1 : The residual of the first two moments (generator vibration acceleration of the generator vibration acceleration sensor); α0: long-term average volatility (determined by historical equipment data statistics of the hydropower unit); α1: weight of the squared residual (reflecting the short-term impact of sudden anomalies on volatility); β1: Conditional variance term weight (reflecting the persistence of volatility, such as long-term fluctuations caused by equipment aging in hydropower units).

[0057] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A green energy fault monitoring system based on a computing network, characterized by: It includes a data acquisition layer, a computing network resource scheduling layer, an intelligent analysis layer, and a fault response layer. The data acquisition layer includes a variety of sensors deployed at each energy node. The algorithm resource scheduling layer adjusts the computing logic of the computing network according to the output power of the green energy power generation network. The data acquisition layer transmits the collected data to the intelligent analysis layer. The intelligent analysis layer outputs the analysis results to the fault response layer, and the fault response layer communicates with the outside world.

2. The green energy fault monitoring system based on computing network according to claim 1 is characterized in that: The green energy power generation grid includes solar power generation units, wind power generation units and hydropower generation units. The solar power generation unit includes output voltage sensor 1, output current sensor 1, solar panel temperature sensor and light intensity sensor. The wind power generation unit includes output voltage sensor 2, output current sensor 2 and wind turbine vibration acceleration sensor. The hydropower generation unit includes output voltage sensor 3, output current sensor 3 and generator vibration acceleration sensor.

3. A green energy fault monitoring method based on a computing network, characterized by: The data collection layer collects the operating data of each part of the green energy power generation network, cleans, standardizes and aggregates the collected data, and selects the system operation mode according to the output power of the green energy power generation network. The operation mode is divided into high green energy mode and low green energy mode. In high green energy mode, tasks are processed according to the size of the dynamic weight of the task. In low green energy mode, except for necessary tasks, all other tasks are migrated to other energy nodes; the intelligent analysis layer calculates the threshold of each node with the data collected by the sensor. The calculation formula for the threshold of each node is: W1=μ1±σ0; σ0=μ1(1-σ1); ； The fault response layer then compares the calculated thresholds of each node with the real-time data of each node, and forms a four-level alarm system, which includes early warning mode, minor alarm mode, severe alarm mode, and disaster mode. When μ1-σ0<W0<μ1+σ0, the system is judged to be in warning level mode, and the operation and maintenance work order is triggered in the warning level mode; When μ1-2σ0<W0<μ1-σ0 or μ1+σ0<W0<μ1+2σ0, the system is judged to be in mild alarm mode. In mild alarm mode, the power output is limited to 80% and the backup heat dissipation is activated; When μ1-3σ0<W0<μ1-2σ0 or μ1+2σ0<W0<μ1+3σ0, the system is judged to be in severe alarm mode, in which emergency shutdown is carried out and fault location information is pushed; When μ1-4σ0<W0<μ1-3σ0 or μ1+3σ0<W0<μ1+4σ0, the system is judged to be in disaster-level mode. In disaster-level mode, the DC side circuit breaker is cut off and the fire protection system is started.

4. The green energy fault monitoring method based on computing network according to claim 3 is characterized in that: Necessary tasks are analytical calculations related to fault monitoring.

5. The green energy fault monitoring method based on computing network according to claim 3 is characterized in that: Normalized data are calculated using the following formula: ; Data aggregation is performed by taking a sliding average of data at multiple time points; In the data cleaning step, the 3-times standard deviation method is used to remove outliers. A data point x is considered an outlier if it meets the following conditions: |x−μ0|>3σ2. The sliding average processing is calculated using the sliding average calculation formula: .

6. The green energy fault monitoring method based on computing network according to claim 3 is characterized in that: The judgment logic of the operation mode is: when P 总 >P 阈值 , the system is in high green energy mode; when P 总 ≤P 阈值 , the system is in low green energy mode.

7. The green energy fault monitoring method based on computing network according to claim 3 is characterized by: Other energy nodes are thermal power generation nodes.

8. The green energy fault monitoring method based on computing network according to claim 3 is characterized by: The calculation formula of the task dynamic weight is: W i =α⋅S i +β⋅(1−U cpu )+γ⋅B n .

9. The green energy fault monitoring method based on computing network according to claim 3 is characterized in that: The network resource scheduling strategies of the network resource scheduling layer are as follows: ① preemptive scheduling: high-weight green energy nodes can preempt tasks of low-weight grid nodes; ② elastic resource reservation: reserve some green energy node resources for sudden fault monitoring tasks.

10. The green energy fault monitoring method based on computing network according to claim 3, characterized in that: Partial green energy node resources account for 25% of the total green energy node resources.

Images