A valve fault diagnosis method for a power plant boiler system

Abstract

The application discloses a valve fault diagnosis method of a power plant boiler system, which comprises the following steps: monitoring points are arranged on the periphery of each valve to acquire monitoring data; the sensors comprise temperature sensors, pressure sensors and flow sensors; all the sensors are connected to a wireless communication module to upload data; a basic database is established; after a diagnosis mode is started, a monitoring period is set, and basic comparison data is screened out; and alarm or safety operation is performed according to the diagnosis result. The application can greatly reduce the requirement of manual inspection by arranging sensors on the periphery of the valve equipment, comparing the diagnosis data with the basic database to judge the fault cause, and giving a fault solution scheme or operation instruction in time, can complete real-time monitoring of the instruments in the factory and rapid processing of faults, and can save a lot of labor cost.

Description

Technical Field

[0001] This invention relates to the field of boiler operation control technology in gas-fired power plants, and specifically to a valve fault diagnosis method for power plant boiler systems. Background Technology

[0002] Power plants utilize various instruments, equipment, and valves. Failure to promptly repair malfunctioning valves can disrupt orderly production and even endanger the lives and property of the entire plant. Traditional valve troubleshooting involves sending an experienced technician for manual inspections. However, manual inspections are inefficient and cannot quickly identify all valve malfunctions, potentially leading to delayed troubleshooting. Furthermore, if maintenance personnel fail to follow safety procedures, it can create threats to personal safety and result in unnecessary losses.

[0003] A power plant boiler, also known as a "power plant boiler," refers to a medium-to-large-sized boiler in a power plant that provides a specified quantity and quality of steam to the steam turbine. It is one of the main thermal equipment in a thermal power plant. It is often paired with a steam turbine generator set of a certain capacity, primarily for power generation, but in some special circumstances, it can also be used for external heating. Generally, it has a large evaporation capacity and high steam parameters (steam temperature and pressure), requiring a complete set of auxiliary equipment. It often needs to be equipped with a combustion furnace, employs forced ventilation, and can burn various fuels (pulverized coal, crude oil or heavy oil, blast furnace gas or coke oven gas). Its structure is relatively complex, and its efficiency is high, often reaching around 85-93%. However, it places considerable demands on its operation and management level, mechanization, and automatic control technology.

[0004] A power plant boiler system typically includes a pulverized coal system, a condensing system, a steam-water system, a turbine unit, a fan system, and a dust removal system. The pulverized coal system includes a coal feeder, a coal mill, and a pulverizer; the condensing system includes a condenser, condensate pumps, a low-pressure heater, a deaerator, a main feedwater pump, a high-pressure heater, and an economizer. The steam-water system includes a steam drum, downcomers, water-cooled walls, a superheater, and a reheater. To enable flow control and start-up / shutdown, the power plant boiler system is equipped with several main valves for corresponding operations. These valves include the first valve between the high-pressure heater and the economizer, the second valve on the downcomer, and the third valve between the reheater and the turbine. During normal operation, the valves generally operate according to preset parameters.

[0005] However, most valves are electronic valves, which facilitate remote monitoring and control. Valves are typically not replaced for many years after installation, and malfunctions can occur. The most common malfunction is that the DCS system controls the valve to open, but due to blockages, signal distortion, errors in the electromagnetic mechanism, or damage, the valve may not open fully, resulting in a certain amount of trapped space. Furthermore, leaks or operational errors during the water supply process can also lead to insufficient water and steam flow in the entire boiler system, ultimately resulting in low boiler efficiency.

[0006] Furthermore, during operation, power plant boilers are often configured with several system operating parameters, namely the power of all equipment; for example, valve flow parameters, fan air volume, pump power, etc. Once these operating parameters are set, each piece of equipment in the power plant boiler system will operate according to these parameters. However, for stability and reliability, the operating parameters of power plant boilers are often relatively fixed, and not a large number of operating modes are set, generally 5-30 operating modes are available. Since each operating mode has corresponding operating parameters, the flow parameters of this invention are not unlimited.

[0007] In summary, current technology relies on manual methods to troubleshoot valve malfunctions in power plants. This approach is inefficient, cannot detect malfunctions in real time, and places high demands on the experience of the troubleshooters, making it prone to errors. Summary of the Invention

[0008] The purpose of this invention is to provide a method for diagnosing valve faults in power plant boiler systems.

[0009] To achieve the above objectives, the present invention provides a valve fault diagnosis method for a power plant boiler system, the power plant boiler system including a pulverized coal system, a condensing gas system, and a steam-water system;

[0010] The pulverized coal system includes a coal feeder, a coal mill, and a pulverizer.

[0011] The condensing system includes a condenser, condensate pumps, low-pressure heaters, a deaerator, a main feedwater pump, a high-pressure heater, and an economizer;

[0012] The steam-water system includes a steam drum, downcomers, water walls, a superheater, and a reheater;

[0013] The power plant boiler system is equipped with several valves, including the first valve between the high-pressure heater and the economizer, the second valve on the downcomer, and the third valve between the reheater and the turbine.

[0014] Valve fault diagnosis methods include the following steps:

[0015] (1) Establish monitoring points:

[0016] Monitoring points are set up around each valve to acquire monitoring data. Sensors are installed at the monitoring points to detect the operating conditions of the fluid medium in the pipeline where each valve is located. The sensors include temperature sensors, pressure sensors, and flow sensors. All sensors are networked and upload data through a wireless communication module.

[0017] (2) Establish a basic database:

[0018] During normal operation of the power plant boiler, the flow data of each valve is acquired through the DCS system, and the monitoring data of the corresponding monitoring point of each valve is acquired through the sensor. The valve flow data and monitoring data at the same time point are correlated to form a set of basic data, and all sets of basic data are used to build a basic database.

[0019] (3) Automatic diagnosis:

[0020] After starting the diagnostic mode, set the monitoring cycle, obtain the flow data of each valve from the power plant's DCS system, as well as the monitoring data of the corresponding monitoring point sensor at the same time, associate the valve flow data and monitoring data at the same time point to form a set of diagnostic data, compare the diagnostic data with each set of basic data in the basic database one by one, and select the basic data that is most similar to the diagnostic data as the basic comparison data.

[0021] (4) Calculate the single approximation value Kd of each valve in the current diagnostic data and the corresponding valve in the baseline comparison data; the single approximation values ​​of the three valves are Kd1, Kd2, and Kd3, respectively. Then calculate the variance value S of each valve, and the variance values ​​of the three valves are S1, S2, and S3, respectively.

[0022] The calculation method for a single approximation is as follows:

[0023]

[0024] Where Kdi is a single approximation of the i-th valve; T is the temperature data in the basic comparison data; N is the flow ratio in the basic comparison data; and P is the pressure data in the basic comparison data.

[0025] The variance of each valve is calculated as follows:

[0026] The largest value among S1, S2 and S3 is taken as the maximum variance value Smax. The maximum variance value Smax is compared with the fault threshold. If the maximum variance value Smax exceeds the fault threshold, the valve corresponding to the maximum variance value Smax is diagnosed as having a fault.

[0027] (5) Issue an alarm or perform a safety operation based on the diagnosis results.

[0028] Preferably, the wireless communication module of this invention is a ZigBee communication module.

[0029] In a preferred embodiment of the present invention, the specific method for selecting the basic data most similar to the diagnostic data as the basic comparison data in the automatic diagnosis step is as follows:

[0030] (31) Calculate the flow ratio of each set of basic data in the basic database, and calculate the flow ratio in each set of diagnostic data. The flow ratio is the flow data set by the valve in the DCS system and the flow monitoring data obtained by the flow sensor at the corresponding monitoring point in each set of data. The flow ratio N = q / Q.

[0031] Where Q is the flow rate data set for the valve in the DCS system at the current time point; q is the flow rate monitoring data obtained by the flow sensor at the monitoring point corresponding to the valve at the current time point;

[0032] (32) Calculate the approximate value K of the diagnostic data related to the three valves and each set of basic data at the current monitoring time point. Any set of diagnostic data and a set of basic data include data of the three valves.

[0033] The formula for calculating the approximate value K is:

[0034]

[0035] Where Ti is the temperature data obtained by the temperature sensor corresponding to the i-th valve in the diagnostic data;

[0036] T0 is the temperature data obtained from the temperature sensor corresponding to the i-th valve in the basic data;

[0037] Ni is the flow ratio corresponding to the i-th valve in the diagnostic data;

[0038] N0 is the flow ratio corresponding to the i-th valve in the basic data;

[0039] Pi is the pressure data obtained by the pressure sensor corresponding to the i-th valve in the diagnostic data;

[0040] P0 is the pressure data obtained from the pressure sensor corresponding to the i-th valve in the basic data;

[0041] The set of basic data with the smallest approximate value K is used as the basic comparison data.

[0042] Preferably, each valve's instrument display is equipped with a camera. During monitoring, the camera acquires the instrument display data of the corresponding valve and associates and stores the image information with diagnostic data. In the event of a fault, the image information is retrieved from the database according to the operation command.

[0043] Preferably, when the first valve is diagnosed as faulty, if the flow ratio is lower than the normal alarm threshold but higher than the severe alarm threshold, an alarm operation command is initiated; if the flow ratio is lower than the severe alarm threshold, the water supply system is activated to replenish water to the deaerator.

[0044] Preferably, when the second valve is diagnosed as faulty, if the flow ratio is lower than the normal alarm threshold but higher than the severe alarm threshold, an alarm operation command is initiated; if the flow ratio is lower than the severe alarm threshold, the pulverized coal system is controlled to operate at reduced load.

[0045] Preferably, in the automatic diagnosis process, the present invention further includes acquiring smoke and gas information of the environment in which the power plant boiler system is located by arranging smoke sensors and gas detection sensors, and associating the information with monitoring data to construct environmental data in the diagnostic data.

[0046] Preferably, when issuing an alarm or performing a safety operation based on the diagnostic results, the present invention acquires the confirmed safety operation records after each fault, forming a fault resolution database.

[0047] In summary, the present invention has the following advantages:

[0048] 1. This invention, by arranging sensors around valve equipment, comparing basic databases with diagnostic data to determine the cause of faults, and promptly providing fault solutions or operating instructions, can greatly reduce the requirements for manual inspections; it can achieve real-time monitoring and rapid fault handling of instruments within the factory; and can save a lot of labor costs.

[0049] 2. This invention uses ZigBee wireless communication to connect with networks, various sensors, cameras and other devices. The database established using the network can greatly avoid situations where professional maintenance personnel are unable to handle faults due to lack of experience, and also greatly ensures personnel safety by eliminating the need for on-site inspections.

[0050] 3. To improve diagnostic accuracy, this invention first searches for the closest set of basic data from a basic database. The closest basic data has a higher similarity to the diagnostic data, allowing for the selection of normal process parameters that most closely match the diagnostic data. These normal process parameters correspond to the basic data; each set of basic data represents a process, or each process forms a set of basic data or several sets of very close approximate basic data. By judging approximations, this invention first filters preset processes close to the process represented by the diagnostic data, ensuring that the compared data all have the same or similar process conditions. This reduces errors caused by differences in the overall process, making the determination of faults more accurate. Furthermore, this invention provides a method for determining the closest approximation, making it easier to accurately identify which valve is malfunctioning. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating a fault diagnosis method in one embodiment of the present invention;

[0052] Figure 2 This is a connection block diagram of the physical system corresponding to the fault diagnosis method in one embodiment of the present invention. Detailed Implementation

[0053] This invention provides a valve fault diagnosis method for a power plant boiler system, which includes a pulverized coal system, a condensing gas system, and a steam-water system. The power plant boiler system described in this invention is existing technology, and the valves addressed in this invention are also installed in existing technology. The modification to the power plant boiler system in this invention lies in the arrangement of sensors and other monitoring equipment around the valves.

[0054] In existing technologies, pulverized coal systems include coal feeders, coal mills, and pulverizers. Coal mills grind lumpy coal into fine powder, while pulverizers mix the fine pulverized coal with preheated air and send it into the boiler's combustion chamber for combustion.

[0055] The condensing system is the main equipment that cools the steam from the turbine and sends it to the steam drum for steam-water circulation. A typical condensing system includes a condenser, condensate pump, low-pressure heater, deaerator, main feedwater pump, high-pressure heater, and economizer. The condenser is connected to the turbine, and the condensate flows sequentially through the condensate pump, low-pressure heater, deaerator, main feedwater pump, high-pressure heater, and economizer before entering the steam drum. In addition, a water supply system is installed to supplement the water source. This system uses river water that has undergone purification and desalination processes to be used as boiler water. The water supply system is directly connected to the deaerator, allowing qualified boiler water to be directly fed into it.

[0056] The steam-water system includes a steam drum, downcomers, water-cooled walls, a superheater, and a reheater. The downcomers allow water from the steam drum to flow into the water-cooled walls, where the water is heated and turned into high-temperature steam. After passing through the superheater and reheater, the steam enters the turbine.

[0057] The power plant boiler system is equipped with several valves, including the first valve between the high-pressure heater and the economizer, the second valve on the downcomer, and the third valve between the reheater and the turbine.

[0058] The valve fault diagnosis method of the present invention includes the following steps:

[0059] (1) Establish monitoring points:

[0060] Monitoring points are set up around each valve to acquire monitoring data. These monitoring points are equipped with sensors to detect the operating conditions of the fluid medium in the pipeline where each valve is located. The sensors include temperature sensors, pressure sensors, and flow sensors. All sensors are networked and transmit data via a wireless communication module. The monitoring points are installed by drilling holes in the pipeline, typically at a distance of about 50 pipe diameters from the valve, to ensure accurate data collection.

[0061] (2) Establish a basic database:

[0062] During normal operation of the power plant boiler, the flow data of each valve is acquired through the DCS system, and the monitoring data of the corresponding monitoring point of each valve is acquired through the sensor. The valve flow data and monitoring data at the same time point are correlated to form a set of basic data, and all sets of basic data are used to build a basic database.

[0063] To enable comparison, baseline data for normal operation needs to be pre-stored in the processor or server. This baseline data refers to accurate data, and its acquisition requires on-site verification by technical personnel to further improve its accuracy. Each operating mode can collect 3-5 sets of baseline data, and the baseline database can store 50-100 sets to ensure usability.

[0064] (3) Automatic diagnosis:

[0065] After activating the diagnostic mode, a monitoring cycle is set, and the flow data of each valve and the monitoring data of the corresponding monitoring point sensor are obtained from the power plant's DCS system. The valve flow data and monitoring data at the same time point are associated to form a set of diagnostic data. The diagnostic data is compared with each set of basic data in the basic database one by one, and the basic data that is most similar to the diagnostic data is selected as the basic comparison data.

[0066] The boiler's operating conditions do not fluctuate drastically, so a monitoring cycle can be set, monitoring every few minutes. This avoids problems and reduces the server's computational load. During operation, the DCS system assigns an on / off status to each valve. This status corresponds to the valve's flow data, meaning the DCS system controls the throughput of each valve during operation. If the DCS system does not change its status, theoretically, the valve will remain in that operating state indefinitely.

[0067] Each set of data in this invention includes valve flow data as well as temperature, flow rate, and pressure data detected by sensors; this allows the valve status and monitoring status at the same point in time to be correlated.

[0068] The diagnostic data of this invention is monitoring data obtained under the current process conditions. By comparing the diagnostic data with the baseline data, and by comparing it with the closest process conditions, the linkage change trends of each valve or device are more consistent when the overall process is similar. Theoretically, when the same process conditions are selected, the data detected by each valve and sensor should be consistent. However, due to the inherent error in monitoring and the instantaneous nature of flow rate changes, there will be some discrepancy between the actual monitored data and the baseline data. If the valves and sensors themselves are not faulty, this difference is often very small. Therefore, this invention determines whether two sets of data are close by calculating approximate values ​​and selects the closest baseline data as the basic comparison data. This can reduce the errors caused by differences in processes.

[0069] In the automatic diagnosis steps of this invention, the specific method for selecting the basic data most similar to the diagnostic data as the basic comparison data is as follows:

[0070] (31) Calculate the flow ratio of each set of basic data in the basic database, and calculate the flow ratio in each set of diagnostic data. The flow ratio is the flow data set by the valve in the DCS system and the flow monitoring data obtained by the flow sensor at the corresponding monitoring point in each set of data. The flow ratio N = q / Q. For example, in the prior art, valves are often more commonly in the fully open state. Therefore, when the valve malfunctions or becomes blocked, the valve is often throttled, that is, the valve is not fully open, and the flow rate is reduced. At this time, the actual monitored flow rate q is lower than the preset flow data set by the valve, resulting in a flow ratio N of less than 1. Conversely, if in some cases the valve malfunctions and the opening degree is greater than the set flow data, the flow ratio N is greater than 1.

[0071] Where Q is the flow rate data set for the valve in the DCS system at the current time point; q is the flow rate monitoring data obtained by the flow sensor at the monitoring point corresponding to the valve at the current time point.

[0072] (32) Calculate the approximate value K of the diagnostic data related to the three valves and each set of basic data at the current monitoring time point. Any set of diagnostic data and a set of basic data include data of the three valves.

[0073] The formula for calculating the approximate value K is:

[0074]

[0075] Where Ti is the temperature data obtained by the temperature sensor corresponding to the i-th valve in the diagnostic data;

[0076] T0 is the temperature data obtained from the temperature sensor corresponding to the i-th valve in the basic data;

[0077] Ni is the flow ratio corresponding to the i-th valve in the diagnostic data;

[0078] N0 is the flow ratio corresponding to the i-th valve in the basic data;

[0079] Pi is the pressure data obtained by the pressure sensor corresponding to the i-th valve in the diagnostic data;

[0080] P0 is the pressure data obtained from the pressure sensor corresponding to the i-th valve in the basic data;

[0081] The set of basic data with the smallest approximate value K is used as the basic comparison data.

[0082] The process pipeline of this invention mainly includes three valves, and the operating conditions of the three valves represent the operation status of the entire process.

[0083] As can be seen from the table above, each set of data includes relevant data for three valves. Each valve includes the valve's set flow rate data, the temperature data monitored by the temperature sensor, the flow rate data monitored by the flow sensor, and the pressure data monitored by the pressure sensor.

[0084] The formula for calculating the approximation value K involves comparing each valve individually with each valve in the baseline comparison data. Specifically, it calculates the approximate value of the first valve in the diagnostic data compared to the first valve in the baseline data, the approximate value of the second valve in the diagnostic data compared to the second valve in the baseline data, and so on. Then, a summation formula is used to calculate the approximate value of this set of diagnostic data compared to the baseline data. The smaller the approximation value, the smaller the difference between the two sets of data, indicating that the two sets of data are more similar and represent more closely related process conditions. Therefore, this invention reduces errors by comparing the diagnostic data with multiple sets of baseline data one by one and selecting the closest baseline data as the comparison data. However, after selecting the baseline comparison data, it is not possible to directly determine whether a valve has a problem, nor can it be determined which valve has a problem. Therefore, this invention uses the variance value of each valve in the following steps to determine which valve has malfunctioned.

[0085] (4) Calculate the single approximation value Kd of each valve in the current diagnostic data and the corresponding valve in the baseline comparison data; the single approximation values ​​of the three valves are Kd1, Kd2, and Kd3, respectively. Then calculate the variance value S of each valve, and the variance values ​​of the three valves are S1, S2, and S3, respectively.

[0086] The calculation method for a single approximation is as follows:

[0087]

[0088] Where Kdi is a single approximation of the i-th valve; T is the temperature data in the basic comparison data; N is the flow ratio in the basic comparison data; and P is the pressure data in the basic comparison data.

[0089] The variance of each valve is calculated as follows:

[0090]

[0091] When calculating the approximate value between each set of diagnostic data and the baseline data, the approximate value K is obtained by summing the three individual approximate values, i.e., K is the sum of Kd1, Kd2, and Kd3. This invention uses the calculation method of variance to screen for valve malfunctions.

[0092] The largest value among S1, S2, and S3 is taken as the maximum variance value Smax. The maximum variance value Smax is compared with a fault threshold. If the maximum variance value Smax exceeds the fault threshold, the valve corresponding to the maximum variance value Smax is diagnosed as faulty. Since monitoring various data cannot completely eliminate errors, this invention sets a fault threshold. Only when the maximum variance value exceeds the threshold can it be determined as a fault; otherwise, it can be considered as an error.

[0093] For example, when a valve malfunctions, its flow ratio and pressure data will change, causing the calculated approximate value to differ from the theoretical value. Conversely, the difference between the calculated approximate value and the theoretical value will be very small for valves that are not malfunctioning. Considering that the operation of the process conditions is a holistic process, directly comparing a single approximate value of a valve with the theoretical value to determine if a valve is malfunctioning does not take into account the process conditions and the conditions of other valves, making it prone to errors. For instance, if the data detected for each valve in the diagnostic data is actually lower than the operating parameters in the baseline comparison data, it indicates that the operating power of the process conditions represented by the diagnostic data is lower than that of the baseline comparison data. Directly comparing the single approximate value of a valve in the diagnostic data with the theoretical value easily leads to errors. Therefore, this invention eliminates this error by using variance analysis.

[0094] Furthermore, due to the instability of monitoring errors or instantaneous monitoring, after detecting a fault in a valve, this invention can immediately collect data from the diagnosed faulty valve 2-5 times for verification. The verification method follows the automatic diagnosis steps to further check whether the valve is faulty. If the verification confirms that the valve is still faulty, it is finally judged as faulty. If the verification confirms that no fault is found, the previous fault determination is cancelled, and the process does not proceed to step (5), and no alarm is triggered.

[0095] (5) Issue an alarm or perform a safety operation based on the diagnosis results.

[0096] Once a problem is diagnosed with a valve, an alarm can be triggered based on preset conditions. This alarm can be displayed on the control panel or triggered by an audible and visual alarm installed on-site or in the monitoring room. When a major valve malfunction is diagnosed, the entire power plant boiler system can be controlled, enabling automated and safe operation.

[0097] For example, when the first valve is diagnosed as faulty, if the flow ratio is below the normal alarm threshold but above the severe alarm threshold, an alarm operation command is initiated. If the flow ratio is below the severe alarm threshold, the water supply system is activated to replenish water to the deaerator. The water supply system can directly supplement the deaerator with purified boiler water. When the second valve is diagnosed as faulty, if the flow ratio is below the normal alarm threshold but above the severe alarm threshold, an alarm operation command is initiated. If the flow ratio is below the severe alarm threshold, the pulverized coal system is controlled to operate at reduced load.

[0098] In the specific implementation of this invention, the wireless communication module is a ZigBee communication module; 2. Communication between sensor devices can be established using the ZigBee wireless module, and the self-organizing network characteristics of ZigBee can basically meet the equipment quantity requirements of most factories, realizing the monitoring and communication tasks of multiple devices.

[0099] In an optimized embodiment of the present invention, a camera can be configured around the instrument display of each valve. During monitoring, the camera acquires the instrument display data of the corresponding valve, and the image information is associated with and stored in conjunction with diagnostic data. In the event of a fault, the image information is retrieved from the database according to the operation command. This allows operators to confirm the valve problem through images when a valve malfunctions, assisting in determining whether a fault has occurred and making a preliminary fault confirmation.

[0100] In an optimized embodiment of the present invention, the automatic diagnostic process further includes acquiring smoke and gas information of the environment in which the power plant boiler system is located by deploying smoke sensors and gas detection sensors, and correlating this information with monitoring data to construct environmental data in the diagnostic data. During plant operation, leaks may occur, which will also affect the flow, temperature, and pressure data monitored by the present invention. Therefore, the present invention further uses the aforementioned sensors to assess the environmental conditions, which helps managers determine the cause and severity of faults in conjunction with environmental factors.

[0101] In an optimized embodiment of the present invention, when an alarm is triggered or a safety operation is performed based on the diagnostic results, a safety operation record confirmed after each fault is obtained to form a fault resolution database. This facilitates the formation of a fault database and makes it easier to provide prompts to operators during subsequent teaching or automated processing.

Claims

1. A valve fault diagnosis method of a power plant boiler system, the power plant boiler system comprising a pulverized coal system, a condensing system and a water-steam system; the pulverized coal system comprising a coal feeder, a coal mill and a pulverized coal feeder; the condensing system comprising a condenser, a condensate pump, a low pressure heater, a deaerator, a main feed water pump, a high pressure heater and an economizer; the water-steam system comprising a steam drum, a downcomer, a water wall, a superheater and a reheater; the power plant boiler system being provided with a plurality of valves, the valves comprising a first valve provided between the high pressure heater and the economizer, a second valve provided on the downcomer, and a third valve provided between the reheater and the steam turbine; the valve fault diagnosis method comprising the following steps: (1) setting monitoring points: setting monitoring points around each valve to obtain monitoring data, the monitoring points being provided with sensors for detecting the working conditions of the fluid medium in the pipeline where each valve is located; the sensors comprising temperature sensors, pressure sensors and flow sensors; all the sensors being connected to a wireless communication module to upload data; (2) establishing a basic database: obtaining flow data of each valve through a DCS system when the power plant boiler is in normal operation, and obtaining monitoring data of the monitoring points corresponding to each valve through the sensors; associating the valve flow data and the monitoring data at the same time point to form a group of basic data, and constructing all groups of basic data into a basic database; (3) automatic diagnosis: after starting the diagnosis mode, setting a monitoring period, obtaining flow data of each valve from the DCS system of the power plant, and obtaining monitoring data of the corresponding sensors at the same time, associating the valve flow data and the monitoring data at the same time point to form a group of diagnosis data, comparing the diagnosis data with each group of basic data in the basic database one by one, and selecting the basic data most similar to the diagnosis data as basic comparison data; (4) calculating a single approximation value Kd of each valve in the current diagnosis data and the corresponding valve in the basic comparison data, the single approximation values of the three valves being Kd1, Kd2 and Kd3 respectively, and then calculating a variance value S of each valve, the variance values of the three valves being S1, S2 and S3 respectively; the calculation method of the single approximation value being: Kd = (q1-q2) / (q1+q2) ; wherein q1 is the flow monitoring data of the corresponding monitoring point of the first valve, q2 is the flow data of the first valve set in the DCS system at the same time point, and q1 and q2 are both positive values; the wireless communication module being a ZigBee communication module; and the specific method of selecting the basic data most similar to the diagnosis data in the automatic diagnosis step being: (31) calculating a flow ratio of each group of basic data in the basic database, and calculating a flow ratio in each group of diagnosis data, the flow ratio being the flow data of the valve set in the DCS system and the flow monitoring data obtained by the flow sensor on the corresponding monitoring point in each group of data, the flow ratio N being q / Q; wherein Q is the flow data of the valve set in the DCS system at the current time point, and q is the flow monitoring data obtained by the flow sensor on the corresponding monitoring point of the valve at the current time point; (32) calculating an approximation value K of the diagnosis data of the three valves related to the current monitoring time point and each group of basic data, any group of diagnosis data and any group of basic data comprising data of the three valves; and the calculation formula of the approximation value K being: K = (q1-q2) / (q1+q2) ; wherein q1 is the flow monitoring data of the corresponding monitoring point of the first valve, q2 is the flow data of the first valve set in the DCS system at the same time point, and q1 and q2 are both positive values. ​ ​ ​ ​ characterized in that ​ ​ ​ ​ ​ ​ ​ ​ ​ Wherein, Kdi is the single approximation value of the ith valve; T is the temperature data in the basic comparison data, Ti is the temperature data monitored by the temperature sensor corresponding to the ith valve in the diagnosis data; N is the flow ratio in the basic comparison data, the flow ratio N=q / Q, wherein Q is the flow data set for the valve in the DCS system at the current time point; q is the flow monitoring data obtained by the flow sensor corresponding to the monitoring point of the valve at the current time point; Ni is the flow ratio corresponding to the ith valve in the diagnosis data; P is the pressure data in the basic comparison data; Pi is the pressure data monitored by the pressure sensor corresponding to the ith valve in the diagnosis data; the calculation method of the variance value of each valve is: The largest value among S1, S2 and S3 is taken as a maximum variance value Smax, the maximum variance value Smax is compared with a fault threshold, if the maximum variance value Smax exceeds the fault threshold, it is diagnosed that the valve corresponding to the maximum variance value Smax is in failure; (5) an alarm is given or a safety operation is performed according to the diagnosis result.

2. The method of claim 1, wherein: ​ 3. The method of claim 1, wherein: ​ ​ ​ ​ wherein; Ti is the temperature data of the i th valve corresponding to the temperature sensor in the diagnostic data; T0 is the temperature data of the i th valve corresponding to the temperature sensor in the basic data; Ni is the flow ratio of the i th valve corresponding to the diagnostic data; N0 is the flow ratio of the i th valve corresponding to the basic data; Pi is the pressure data of the i th valve corresponding to the pressure sensor in the diagnostic data; P0 is the pressure data of the i th valve corresponding to the pressure sensor in the basic data; the group of basic data with the smallest approximation value K is taken as the basic comparison data. The periphery of the instrument display of each valve is configured with a camera, and the instrument display data of the corresponding valve is acquired through the camera during monitoring, and the image information corresponding to the instrument display data is stored in association with the diagnostic data, and the image information is called from the database according to the operation instruction when a fault occurs.

4. The method of claim 1, wherein: When the first valve is diagnosed as a fault, if the flow ratio is lower than the ordinary alarm threshold and greater than the serious alarm threshold at this time, the alarm operation instruction is started, and if the flow ratio is lower than the serious alarm threshold at this time, the water source water replenishing system is started to replenish water to the deaerator.

5. The method of claim 1, wherein: When the second valve is diagnosed as a fault, if the flow ratio is lower than the ordinary alarm threshold and greater than the serious alarm threshold at this time, the alarm operation instruction is started, and if the flow ratio is lower than the serious alarm threshold at this time, the pulverized coal system is controlled to run at a reduced load.

6. The method of claim 1, wherein: In the automatic diagnosis process, the smoke and gas information of the environment of the power plant boiler system is acquired through the arrangement of smoke sensors and gas detection sensors, the information is associated with the monitoring data, and the environment data in the diagnostic data is constructed.

7. The method of claim 1, wherein: When the alarm or safety operation is performed according to the diagnostic result, the safety operation record confirmed after each fault is acquired, and a fault solution database is formed.

8. The method of claim 1, wherein: ​

Images