Water injection well diagnosis and early warning method based on node analysis and multi-information fusion technology

Abstract

The invention discloses a water injection well diagnosis and early warning method based on node analysis and a multi-information fusion technology. The method comprises the following steps that node analysis is conducted on a water injection system of an oil field; performing multi-information fusion on the basic data of the water injection well and the node analysis result, and diagnosing the operation condition of the water injection well in real time to obtain a condition diagnosis result; and the working condition diagnosis result is pushed and processed in time, and the processing progress is detected or the processing result is fed back. According to the water injection well diagnosis and early warning method based on the node analysis and the multi-information fusion technology, the node analysis technology and the multi-information fusion technology are utilized, working condition judgment is refined, many factors are considered, historical working conditions are tracked in time, meanwhile, real-time judgment, early warning in advance and timely pushing processing can be achieved, and continuous tracking is achieved; water injection well working condition real-time diagnosis early warning is achieved, and the water injection well management level is further improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of water supply in the oil industry, and particularly relates to a method for diagnosing and warning injection wells based on node analysis and multi-information fusion technology. Background Art

[0002] Most domestic oil fields need to supplement energy to the formation through artificial measures to maintain the formation pressure, achieve stable oil production in the oil field, and improve the ultimate recovery rate of the oil field. Among them, water injection is an effective means and the most important measure to maintain the formation pressure of the reservoir and ensure the efficient and economic development of the reservoir. Whether the injection well operates normally seriously affects the water injection development effect. How to efficiently diagnose the working conditions of injection wells and give early warnings to ensure the normal operation of water injection is an important part of the production management of injection wells.

[0003] Fault diagnosis is to analyze and process the information obtained by testing the diagnostic object using various instrument sensors and detection methods, judge whether the state of the diagnostic object is in an abnormal state or a fault state, and then determine the type of the fault, and perform fault location, tracking and warning.

[0004] With the rapid development of electronic computers, the Internet, big data and artificial intelligence technologies, the existing fault diagnosis of injection wells has poor timeliness and low diagnostic accuracy, which easily leads to lag in fault analysis and lack of early warning tracking. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for diagnosing and warning injection wells based on node analysis and multi-information fusion technology, which solves the problems of poor timeliness of fault diagnosis of injection wells and low diagnostic accuracy in the existing technology, resulting in lag in fault analysis and lack of early warning tracking.

[0006] The technical solution adopted by the present invention is a method for diagnosing and warning injection wells based on node analysis and multi-information fusion technology, which is specifically implemented according to the following steps:

[0007] Step 1: Perform node analysis on the water injection system of the oil field;

[0008] Step 2: Perform multi-information fusion on the basic data, dynamic real-time data of the injection well and the node analysis result of Step 1, and diagnose the operation condition of the injection well in real time to obtain the operation condition diagnosis result;

[0009] Step 3: Push and process the operation condition diagnosis result of Step 2 in a timely manner, and monitor and feedback the processing progress.

[0010] The characteristics of the present invention also lie in:

[0011] Step 1 is specifically as follows:

[0012] The water injection system is divided into two subsystems with the steady flow valve as the node: the wellbore system and the surface water injection system. The wellbore system includes three nodes: the wellhead, the bottom of the well, and the formation. The surface water injection system includes the water source well, the water injection station, the high-pressure valve group room, and the network node. Network monitoring equipment, control equipment, flow meters, and pressure gauges are installed at each node of the entire water injection system to conduct node analysis on the wellbore system.

[0013] The node analysis is specifically divided into three steps:

[0014] Step 1.1: Draw the inflow performance curve of the node;

[0015] Specifically: Calculate the pressure values of each node at different flow rates to obtain the pressure distribution, and draw the inflow performance curve based on the pressure and flow rate results;

[0016] Among them, the wellhead pressure data is the real-time monitoring data of the network monitoring equipment and instruments. The bottom hole pressure value is calculated as follows: Taking the bottom hole as the solution point, starting from the tubing pressure of each well in the valve group, calculate the bottom hole pressure values of the same well at different flow rates respectively;

[0017] Step 1.2: Draw the bottom hole IPR curve at different water injection rates;

[0018] Specifically: Calculate the formation water absorption starting pressure and water absorption index value and draw the bottom hole IPR curve;

[0019] Among them, the water absorption index value is the ratio of the daily injection volume of the water injection well to the wellhead pressure. The formation water absorption starting pressure: the injection pressure when the injection volume is zero, and the formation water absorption starting pressure is obtained by extending the water absorption curve;

[0020] Step 1.3: Draw the inflow performance curve in Step 1.1 and the bottom hole IPR curve in Step 1.2 in the same coordinate system to obtain the node analysis calculation results. The node analysis calculation results include the node flow law, pressure distribution, and supply-injection balance point. The supply-injection balance point is the intersection point of the two curves.

[0021] In Step 1.2, the formation water absorption starting pressure and water absorption index value can also be obtained according to Darcy's law of permeability and the bottom hole IPR curve is drawn.

[0022] In Step 2, the basic data includes water injection dynamic and static data, surrounding oil well dynamic data, physical property parameters, fracture orientation, wellbore water absorption situation, and wellhead status.

[0023] In Step 2, the multi-information fusion is specifically as follows: Use the random forest algorithm for data classification, learning, and judgment to real-time diagnose six operating conditions: water injection well network warning, station control warning, instrument warning, over-injection and under-injection warning, pressure warning, and wellbore diagnosis warning.

[0024] The six operating condition diagnosis conditions are as follows:

[0025] 1) Network warning: The network monitoring devices at the water source wells, water supply and injection stations, high-pressure valve groups, and network nodes in the surface water injection system regularly send messages to the server. The server analyzes the network operation status of each node and sends the results to the database, comprehensively analyzes the network situation. If the server fails to receive the socket message within the timeout period, it is regarded as abnormal network, and then it is determined as a network warning. The server reports the network warning, stores the warning result in the database, and displays the warning result on the server system page; otherwise, it is judged as normal;

[0026] 2) Station control warning: Check the link status of the real-time points of the control devices at each node in the surface water injection system. If the statuses of the four real-time points of pipe pressure, manifold pressure, instantaneous flow rate, and cumulative flow rate are all that data has not been received due to timeout, the server reports a station control warning and stores the result in the database; otherwise, it is judged as normal;

[0027] 3) Instrument warning: Check the cumulative flow rate value of each flowmeter head in the surface water injection system. If the cumulative flow rate value is less than 0 or the change of the head exceeds 2 times the well water injection allocation within five minutes, the server reports a cumulative head failure; check the manifold pressure value. If the manifold pressure value is less than 0, greater than 2 times the range, or the last three or four digits after the decimal point of the data do not change for a long time, the server reports a manifold pressure gauge head failure; check the pipe pressure. If the pipe pressure is less than 0, exceeds 2 times the range, or the last three or four digits after the decimal point of the data do not change for a long time, the server reports a pipe pressure gauge head failure; otherwise, it is judged as normal;

[0028] 4) Over-injection and under-injection warning: For injection wells with a daily injection allocation of less than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection allocation to the daily injection allocation is outside the range of -15% to 15%, the server reports an over-injection and under-injection warning; otherwise, it is judged as normal; for injection wells with a daily injection allocation of more than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection allocation to the daily injection allocation is outside the range of -10% to 10%, the server reports an over-injection and under-injection warning; otherwise, it is judged as normal;

[0029] 5) Pressure warning: For the surface water injection system and the injection wellbore system, if the manifold pressure, pipe pressure, oil pressure, and casing pressure on the same pipeline do not meet the condition of pressure ≥ pipe pressure ≥ oil pressure ≥ casing pressure, the server reports a pressure warning; otherwise, it is judged as normal;

[0030] 6) Wellbore diagnosis and early warning: For the surface water injection system and the injection wellbore system, if the oil pressure drops by more than the set range compared with the previous day, the server reports an abnormal wellbore early warning; if the difference between the oil pressure and the casing pressure decreases by more than the set range or is 0 compared with the previous day, the server reports a packer abnormal early warning; if the oil pressure rises by more than 20% compared with the previous day, the actual water injection volume decreases, and an under-injection fault occurs, the server reports a wellbore blockage early warning, otherwise it is judged normal.

[0031] Step 3 is specifically as follows: Set the processing time limit, use human-computer interaction to immediately feedback the working condition diagnosis result of Step 2 to the trigger, and the trigger pushes the result to the relevant staff. If it is not processed beyond the processing time limit, the trigger is activated, and the result is fed back to the higher-level personnel, and so on until the problem is solved.

[0032] The beneficial effects of the present invention are:

[0033] The wellbore diagnosis and early warning method based on node analysis and multi-information fusion technology of the present invention uses node analysis technology and multi-information fusion technology, with refined working condition determination, considering many factors, tracking historical working conditions in a timely manner, being able to make real-time judgments, giving early warnings and pushing for processing in a timely manner, and continuously tracking, realizing real-time diagnosis and early warning of the working conditions of injection wells, further improving the management level of injection wells, and at the same time reducing the number of on-site inspections by personnel and reducing the labor intensity. Description of the Drawings

[0034] Figure 1 is the flowchart of node analysis in the method of the present invention;

[0035] Figure 2 is the block diagram of the node analysis calculation process in the method of the present invention;

[0036] Figure 3 is the flowchart of the multi-information fusion technology in the method of the present invention;

[0037] Figure 4 is the block diagram of the multi-information fusion model and the working condition early warning process in the method of the present invention;

[0038] Figure 5 is the block diagram of the fault early warning push and processing process in the method of the present invention;

[0039] Figure 6 is the overall process schematic diagram of the wellbore diagnosis and early warning method of the present invention based on node analysis and multi-information fusion technology. Detailed Embodiments

[0040] The present invention will be described in detail below with reference to the drawings and specific embodiments.

[0041] Embodiment 1

[0042] The injection well diagnosis and early warning method based on node analysis and multi-information fusion technology has a process as follows Figure 6 shown, and is specifically implemented according to the following steps:

[0043] Step 1: Conduct node analysis on the injection system of the oilfield;

[0044] Divide the injection system into two subsystems with the steady flow valve as the node: the injection wellbore system and the surface injection system. As Figure 1 shown, the injection wellbore system includes three nodes: the injection wellhead, the injection well bottom, and the formation. The surface injection system includes four nodes: the water source well, the water injection station, the high-pressure valve group room, and the network node. Network monitoring equipment, control equipment, flow meters, and pressure gauges are installed at each node of the entire injection system. For the injection wellbore system, use the known and calculated injection data to carry out node analysis of the injection well.

[0045] The surface injection system mainly obtains real-time data from the installed network monitoring equipment and instruments. The injection wellbore system makes calculations based on the measured data.

[0046] Step 2: Multi-information fusion diagnosis technology, the process is as Figure 3 and Figure 4 shown. Take the dynamic and static injection data, dynamic data of surrounding oil wells, physical property parameters, fracture orientation, wellbore water absorption conditions, wellhead status, etc. as basic data, and combine the dynamic real-time data of the injection well and the node analysis calculation results obtained in Step 1 to diagnose the operating conditions of the injection well in real time and obtain the condition diagnosis results;

[0047] Use the random forest algorithm to classify, learn, and judge the basic data of the injection well, the dynamic real-time data, and the node analysis results in Step 1, diagnose the operating conditions of the injection well in real time, and make the recognition results richer and more accurate. As Figure 3 and Figure 4 shown, the historical data of the injection well includes the basic data of the injection well and the node analysis result data in Step 1;

[0048] Step 3: Human-machine interaction mode to promote timely handling of conditions. Set a processing time limit, and use human-machine interaction to timely feedback the condition diagnosis results in Step 2 to the trigger. The trigger will push the results to relevant staff through enterprise WeChat or instant messaging. If it is not processed beyond the processing time limit, the trigger will be activated and the results will be fed back to the higher-level personnel, and so on until the problem is solved. As Figure 5 and Figure 6 shown. For the condition diagnosis results obtained in Step 2, push the results, realize human-machine interaction, conduct on-site verification, comparison, and verification, and feedback the results to the system after on-site investigation, which is convenient for the model to self-learn and further optimize, and further correct the diagnosis model.

[0049] Example 2

[0050] Based on Example 1, the node analysis is specifically as Figure 2 shown, which is divided into three steps:

[0051] Step 1.1: Draw the inflow curve of the node;

[0052] Specifically: Calculate the pressure values of each node at different flow rates to obtain the pressure distribution of the injection wellbore system, and plot the pressure and flow rate results into an inflow performance curve;

[0053] Among them, the injection wellhead pressure data is obtained from the real-time data of the installed network monitoring equipment and instruments; the bottom hole pressure value is calculated as: taking the bottom hole as the solution point, starting from the tubing pressure of each well in the valve group, calculate the bottom hole pressure values at different flow rates of the same well respectively.

[0054] Step 1.2: Draw the bottom hole IPR curve at different injection volumes;

[0055] Specifically: Calculate the formation water absorption starting pressure and water absorption index value and draw the bottom hole IPR curve. Among them, the water absorption index value is the ratio of the daily injection volume of the injection well to the wellhead pressure, which is calculated. Formation water absorption starting pressure: The injection pressure when the injection volume is zero, that is, at the moment of injection, through the extension of the water absorption curve, the formation water absorption starting pressure can be obtained;

[0056] Step 1.3: Plot the two curves of the inflow curve in Step 1.1 and the bottom hole IPR curve in Step 1.2 in the same coordinate system to obtain the node analysis calculation results. The node analysis calculation results include the node flow law, pressure distribution, and injection-production balance point. The injection-production balance point is the intersection point of the two curves, also known as the injection volume coordination point. At the intersection point, the inflow and outflow amounts are equal, achieving injection-production balance.

[0057] The node analysis method can realize the analysis and monitoring of the entire injection well system. Through the node analysis calculation results, the injection well can reach the target injection volume under the production requirements, thereby improving production efficiency.

[0058] Example 3

[0059] Based on Example 2, in Step 1.2, the formation water absorption starting pressure and water absorption index value can also be calculated according to Darcy's permeability formula and the bottom hole IPR curve can be drawn.

[0060] Example 4

[0061] Based on Example 3, in Step 2, six types of operating condition warnings that can be analyzed for real-time diagnosis of the injection well operation conditions are: network warning, station control warning, instrument warning, over-injection / under-injection warning, pressure warning, wellbore diagnosis warning, etc. The analysis conditions for each warning are:

[0062] ① Network warning: The network monitoring devices at water source wells, water injection stations, high-pressure valve groups, and network nodes regularly send socket reports to the SCADA server. The SCADA server analyzes the network operation status of 7 common points at each node and sends the results to the database. If the server fails to receive the socket report after a timeout, it is regarded as abnormal network, and a network warning is determined. The SCADA server reports the network warning, stores the warning result in the database, and displays the warning result on the SCADA system page; otherwise, it is judged as normal network.

[0063] ② Station control warning: Check the link status of the real-time points of the control devices at each node in the surface water injection system. If the statuses of the four real-time points of pipe pressure, manifold pressure, instantaneous flow rate, and cumulative flow rate are all in timeout status, that is, there is no data, the SCADA server reports the station control warning result and stores the result in the database; otherwise, it is judged as normal.

[0064] ③ Instrument warning: As Figure 3 shown, check the cumulative flow rate value of each flowmeter head in the surface water injection system. If the cumulative flow rate value is less than 0 or the change of the head exceeds 2 times the well water injection allocation within five minutes, the cumulative head fails; check the manifold pressure value. If the manifold pressure value is less than 0, greater than 2 times the range, or the decimal places after the data do not change for a long time, the manifold pressure gauge head fails; check the pipe pressure. If the pipe pressure is less than 0, exceeds 2 times the range, or the decimal places after the data do not change for a long time, the pipe pressure gauge head fails; otherwise, it is judged as normal.

[0065] ④ Over-injection / under-injection warning: For water injection wells with a daily injection allocation of less than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection allocation to the daily injection allocation is < -15% or > 15%, the server reports an over-injection / under-injection warning; for water injection wells with a daily injection allocation of more than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection allocation to the daily injection allocation is < -10% or > 10%, the server reports an over-injection / under-injection warning.

[0066] ⑤ Pressure warning: For the surface water injection system and the wellbore system of water injection wells, if the manifold pressure, pipe pressure, oil pressure, and casing pressure on the same pipeline do not meet the condition of manifold pressure ≥ pipe pressure ≥ oil pressure ≥ casing pressure, and the pressure fluctuation exceeds the set range, the server reports a pressure warning; otherwise, it is judged as normal.

[0067] ⑥Wellbore diagnosis and early warning: For the surface water injection system and the injection wellbore system, compared with the previous day, if the oil pressure drops by more than the set range, that is, the oil pressure is greater than the casing pressure, and their difference cannot be negative. Generally, a data is given every 5 minutes. If there is a sudden drop, that is, the data drops discontinuously, then the wellbore is abnormal; compared with the previous day, if the difference between the oil pressure and the casing pressure decreases by more than the normal range or is 0, then the packer is abnormal; compared with the previous day, if the oil pressure rises by more than 20% and the actual water injection volume decreases and an under-injection fault occurs, then the wellbore is blocked.

[0068] The working principle of the injection well diagnosis and early warning method based on node analysis and multi-information fusion technology of the present invention is as follows: By using node analysis technology and multi-information fusion technology, the working condition determination is refined, with many factors considered, the historical working conditions are tracked in a timely manner, and at the same time, real-time determination can be achieved, early warning can be pushed in a timely manner for processing, and continuous tracking is carried out to realize real-time diagnosis and early warning of the injection well working conditions, and further improve the management level of injection wells.

[0069] The advantages of the injection well diagnosis and early warning method based on node analysis and multi-information fusion technology of the present invention are as follows: The node analysis is finer, more factors are considered, and historical early warning data can also be combined, so the results obtained by early warning are more accurate; Real-time early warning can be achieved, and the timeliness is stronger; The working condition results are pushed in a timely manner and the processing time limit is set to detect and feedback the processing progress, so that the processing and feedback results can be obtained more timely, which is convenient for adjustment; The timeliness of the injection well working condition processing is stronger, ensuring sufficient and good water injection, improving the water injection effect, ensuring the normal operation of water injection, reducing the number of on-site inspections by personnel at the same time, and reducing the labor intensity.

Claims

1. A method for diagnosing and warning water injection wells based on node analysis and multi-information fusion technology, characterized in that, The implementation is specifically carried out according to the following steps: Step 1: Conduct node analysis on the water injection system of the oilfield; Step 2: Perform multi-information fusion on the basic data, dynamic real-time data of the injection well, and the node analysis results of Step 1, diagnose the operating conditions of the injection well in real time, and obtain the condition diagnosis results; Step 3: Push and process the condition diagnosis results of Step 2 in a timely manner, monitor and feedback the processing progress.

2. The water injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 1, wherein The specific content of Step 1 is as follows: The water injection system is divided into two subsystems with the steady flow valve as the node: the injection wellbore system and the surface water injection system. The injection wellbore system includes three nodes: the injection wellhead, the injection well bottom, and the formation. The surface water injection system includes the water source well, the water injection station, the high-pressure valve group room, and the network node. Network monitoring equipment, control equipment, flow meters, and pressure gauges are installed at each node of the entire water injection system to conduct node analysis on the injection wellbore system.

3. The water injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 2, characterized in that The node analysis is specifically divided into three steps: Step 1.1: Draw the inflow performance curve of the node; Specifically: Calculate the pressure values of each node at different flow rates to obtain the pressure distribution, and draw the inflow performance curve according to the pressure and flow results; Among them, the injection wellhead pressure data is the data real-time monitored by the network monitoring equipment and instruments. The bottom hole pressure value is calculated as follows: Taking the bottom hole as the solution point, starting from the pipe pressure of each well in the valve group, calculate the bottom hole pressure values of the same well at different flow rates respectively; Step 1.2: Draw the bottom hole IPR curve at different water injection rates; Specifically: Calculate the formation water absorption starting pressure and the water absorption index value and draw the bottom hole IPR curve; Among them, the water absorption index value is the ratio of the daily injection volume of the injection well to the wellhead pressure. The formation water absorption starting pressure: the injection pressure when the injection volume is zero, and the formation water absorption starting pressure is obtained by extending the water absorption curve; Step 1.3: Draw the inflow performance curve of Step 1.1 and the bottom hole IPR curve of Step 1.2 in the same coordinate system to obtain the node analysis calculation results. The node analysis calculation results include the node flow law, pressure distribution, and injection-production balance point. The injection-production balance point is the intersection point of the two curves.

4. The injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 3, wherein, In Step 1.2, the formation water absorption starting pressure and the water absorption index value can also be obtained according to Darcy's penetration formula and the bottom hole IPR curve is drawn.

5. The well injection diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 3, characterized in that In Step 2, the basic data includes the dynamic and static water injection data, the dynamic data of surrounding oil wells, physical property parameters, fracture orientation, wellbore water absorption situation, and wellhead status.

6. The water injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 5, characterized in that, In Step 2, the multi-information fusion is specifically as follows: Use the random forest algorithm for data classification, learning, and judgment to diagnose six operating conditions of the injection well network warning, station control warning, instrument warning, over-injection and under-injection warning, pressure warning, and wellbore diagnosis warning in real time.

7. The injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 6, characterized in that The diagnosis conditions of the six operating conditions are respectively: 1) Network warning: The network monitoring devices at the water source wells, water injection stations, high-pressure valve groups, and network nodes in the surface water injection system send messages to the server at regular intervals. The server analyzes the network operation status of each node and sends the results to the database, comprehensively analyzes the network situation. If the server fails to receive the socket message within the timeout period, it is regarded as abnormal network, and then a network warning is determined. The server reports the network warning, stores the warning result in the database, and displays the warning result on the server system page; Otherwise, it is judged as normal; 2) Station control warning: Check the link status of the real-time points of the control devices at each node in the surface water injection system. If the status of the four real-time points of pipeline pressure, manifold pressure, instantaneous flow rate, and cumulative flow rate all show that data has not been received within the timeout period, the server reports a station control warning and stores the result in the database. Otherwise, it is judged as normal; 3) Instrument warning: Check the cumulative flow rate value of each flowmeter head in the surface water injection system. If the cumulative flow rate value is less than 0 or the change of the head exceeds twice the water injection volume of the water well within five minutes, the server reports a cumulative head failure; Check the manifold pressure value. If the manifold pressure value is less than 0, greater than twice the range, or the third and fourth digits after the decimal point of the data do not change for a long time, the server reports a manifold pressure gauge head failure; Check the pipeline pressure. If the pipeline pressure is less than 0, exceeds twice the range, or the third and fourth digits after the decimal point of the data do not change for a long time, the server reports a pipeline pressure gauge head failure. Otherwise, it is judged as normal; 4) Over-injection / under-injection warning: For water injection wells with a daily injection volume of less than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection volume to the daily injection volume is outside the range of -15% to 15%, the server reports an over-injection / under-injection warning. Otherwise, it is judged as normal; For water injection wells with a daily injection volume of more than 20 cubic meters, if the ratio of the difference between the actual injection volume and the daily injection volume to the daily injection volume is outside the range of -10% to 10%, the server reports an over-injection / under-injection warning; Otherwise, it is judged as normal; 5) Pressure warning: For the surface water injection system and the wellbore system of the water injection well, if the manifold pressure, pipeline pressure, oil pressure, and casing pressure on the same pipeline do not meet the requirement of pressure ≥ pipeline pressure ≥ oil pressure ≥ casing pressure, the server reports a pressure warning. Otherwise, it is judged as normal; 6) Wellbore diagnosis warning: For the surface water injection system and the wellbore system of the water injection well, compared with the previous day, if the oil pressure drops by more than the set range, the server reports a wellbore abnormality warning; Compared with the previous day, if the difference between the oil pressure and the casing pressure decreases by more than the set range or is 0, the server reports a packer abnormality warning; Compared with the previous day, if the oil pressure rises by more than 20%, and the actual injection volume decreases and an under-injection fault occurs, the server reports a wellbore blockage warning. Otherwise, it is judged as normal.

8. The water injection well diagnosis and early warning method based on node analysis and multi-information fusion technology according to claim 7, characterized in that, The specific steps of step 3 are as follows: Set the processing time limit, use human-computer interaction to immediately feedback the working condition diagnosis result of step 2 to the trigger, and the trigger pushes the result to the relevant staff. If it is not processed beyond the processing time limit, the trigger is activated, and the result is feedback to the higher-level personnel, and so on until the problem is solved.

Images