A method, medium and system for estimating the health of a converter valve

Abstract

The present invention discloses a method for estimating the health of a converter valve, comprising: obtaining monitoring data of the converter valve; estimating the health of the converter valve according to the size relationship between the monitoring data and protection data and rated data. The embodiment of the present invention designs different judgment logics and uses multi-dimensional data to estimate the health of the converter valve, which is highly efficient and is conducive to timely discovering potential abnormalities of the converter valve for prevention.

Description

Technical Field

[0001] The present invention relates to the technical field of converter valves, and in particular to a method, medium and system for estimating the health of a converter valve. Background Art

[0002] The converter valve is the core equipment of UHV DC transmission. The failure of the converter valve will not only lead to the shutdown of DC transmission, causing serious economic losses, but also cause major safety accidents. If the temperature, pressure, abnormal discharge and working status of the key components of the converter valve can be effectively monitored, combined with the holographic perception and health status assessment and intelligent diagnosis technology of the converter valve, the safety hazards of the equipment can be discovered and eliminated in time, which can effectively improve the reliability of the converter valve operation, save operation and maintenance costs, reduce unnecessary economic losses, and avoid safety accidents.

[0003] At present, domestic and foreign researchers can only make logical judgments on whether the converter valve is in a normal state, but cannot judge the health status of the converter valve, let alone predict the expected life and maintenance and replacement plans of the key components of the converter valve. The operation and maintenance of the converter valve is an important technical means to ensure its safety and reliability. In addition, during the operation and inspection of the converter valve, the staff lacks estimation and analysis of the health status of the valve. Many tasks are cause analysis and treatment after the problem occurs, and there is a lack of clear goals and corresponding means to maintain and repair the equipment. Summary of the invention

[0004] The embodiments of the present invention provide a method, a medium and a system for estimating the health status of a converter valve, so as to solve the problem that the prior art cannot determine the health status of the converter valve.

[0005] In a first aspect, a method for estimating the health of a converter valve is provided, comprising:

[0006] Obtain monitoring data of the converter valve;

[0007] The health of the converter valve is estimated according to the relationship between the monitoring data, the protection data and the rated data.

[0008] In a second aspect, a computer-readable storage medium is provided, on which computer program instructions are stored; when the computer program instructions are executed by a processor, a method for estimating the health of a converter valve as described in the embodiment of the first aspect is implemented.

[0009] In a third aspect, a system for estimating the health of a converter valve is provided, comprising: a computer-readable storage medium as described in the embodiment of the second aspect above.

[0010] Thus, in the embodiment of the present invention, different judgment logics are designed to use multi-dimensional data to estimate the health of the converter valve, which is highly efficient and helps to timely discover potential abnormalities of the converter valve for prevention. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0012] Figure 1 is a flow chart of a method for estimating the health status of a converter valve according to Embodiment 1 of the present invention;

[0013] Figure 2 is a flow chart of a method for estimating the health status of a converter valve according to Embodiment 2 of the present invention;

[0014] Figure 3 This is a flow chart of a method for estimating the health status of a converter valve according to Embodiment 3 of the present invention;

[0015] Figure 4 This is a flow chart of a method for estimating the health status of a converter valve according to Embodiment 4 of the present invention;

[0016] Figure 5 This is a flow chart of a method for estimating the health of a converter valve according to Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Example 1

[0019] Embodiment 1 of the present invention discloses a method for estimating the health of a converter valve. The converter valve is generally a DC converter valve. Figure 1 As shown, the method comprises the following steps:

[0020] Step S101: Acquire monitoring data of the converter valve.

[0021] The type of monitoring data can be selected according to the actual situation. For example, the types of monitoring data include: voltage signal, current signal, power signal, temperature signal, humidity signal, abnormal discharge signal. The monitoring data can be within a preset time range.

[0022] Step S102: Estimate the health of the converter valve according to the relationship between the monitoring data, the protection data and the rated data.

[0023] The protection data can be obtained by preset, which represents the data when the protection device of UHV DC transmission is in action. The rated data is the preset normal data.

[0024] The types of protection data and rated data are the same as those of monitoring data. When comparing the size relationship, the data of the same type are compared.

[0025] For example, the types of protection data also include voltage signals, current signals, power signals, temperature signals, humidity signals, and abnormal discharge signals. The types of rated data also include voltage signals, current signals, power signals, temperature signals, humidity signals, and abnormal discharge signals.

[0026] Embodiment 1 of the present invention can estimate the health of the converter valve by monitoring the relationship between the data, the protection data and the rated data.

[0027] Example 2

[0028] Embodiment 2 of the present invention discloses a method for estimating the health of a converter valve. The converter valve is generally a DC converter valve. Figure 2 As shown, the method comprises the following steps:

[0029] Step S201: Collecting real-time monitoring data of the converter valve.

[0030] The type of real-time monitoring data can be selected according to actual conditions. For example, the types of real-time monitoring data may include: voltage signal, current signal, power signal, temperature signal, humidity signal, abnormal discharge signal. Real-time monitoring data can be collected through corresponding sensors. For example, different types of structured real-time monitoring data can be collected through voltage sensors, current sensors, temperature sensors, humidity sensors, etc. These real-time monitoring data can be collected at preset intervals and can be collected within a preset time range. The data directly collected by the processor is primary data, which can be normalized. The main normalization processing includes filtering, eliminating noise, extracting fixed time intervals, etc., so that the processed real-time monitoring data is more convenient for application in subsequent steps.

[0031] In some cases, real-time monitoring data also includes semi-structured data. These semi-structured data are obtained based on audio, video and photo shooting devices. Semi-structured data can be uniformly adjusted in clarity, brightness and chromaticity to generate normalized semi-structured data. Semi-structured data can be used together with the aforementioned structured data in subsequent steps, or can be used alone in subsequent steps.

[0032] Processing and classifying data can make the data more standardized, more efficient and more stable.

[0033] Step S202: If the real-time monitoring data is closer to the protection data, it is estimated that the health of the converter valve is low.

[0034] Step S203: If the real-time monitoring data is closer to the rated data, the health of the converter valve is estimated to be high.

[0035] For step S202 and step S203, the same type of real-time monitoring data is compared with the protection data and the rated data.

[0036] In addition, the static health margin data corresponding to each real-time monitoring data can be further outputted according to the static margin data of the real-time monitoring data between the protection data and the rated data.

[0037] Embodiment 2 of the present invention can estimate the health of the converter valve according to the degree of proximity between the real-time monitoring data and the protection data and the rated data.

[0038] Example 3

[0039] Embodiment 3 of the present invention discloses a method for estimating the health of a converter valve. The converter valve is generally a DC converter valve. Figure 3 As shown, the method comprises the following steps:

[0040] Step S301: Collecting real-time monitoring data of the converter valve.

[0041] This step is the same as step S201 and will not be described again here.

[0042] Step S302: Count the first amount of real-time monitoring data that exceeds the range defined by the protection data and the rated data.

[0043] Similarly, when taking statistics, the same type of real-time monitoring data, protection data and rated data are also counted.

[0044] Specifically, the first quantity is calculated using the following calculation formula:

[0045] D=count[(S i -B i )|(E i-S i )].

[0046] Where D represents the first quantity, S i represents the i-th real-time monitoring data, B i represents the protection data corresponding to the i-th real-time monitoring data, E i represents the rated data corresponding to the i-th real-time monitoring data, and the vertical line represents or. Therefore, this formula represents S i With B i The difference and E i With S i If one of the differences does not meet the requirements, that is, the real-time monitoring data exceeds the range limited by the protection data and the rated data, it will be counted once.

[0047] Step S303: Calculate a first ratio of the first quantity to the total quantity of all real-time monitoring data.

[0048] The calculation ratio is also calculated for the same type. All real-time monitoring data is all real-time monitoring data collected within the first preset time range.

[0049] Step S304: estimating a first health level of the converter valve according to the first ratio.

[0050] The first health level can be preset. For example, the first health level can be divided into healthy, slightly unhealthy, and severely unhealthy. Each first health level corresponds to a numerical range. According to the numerical range in which the first ratio is located, the first health level is estimated.

[0051] Specifically, the following calculation formula is used to estimate the first health level:

[0052]

[0053] Wherein, J1 is the first health level, D1 is the first ratio, J1=2 corresponds to severe unhealth, J1=1 corresponds to mild unhealth, and J1=0 corresponds to health.

[0054] Embodiment 3 of the present invention can determine the corresponding health level by the proportion of real-time monitoring data that exceeds the range defined by the protection data and the rated data.

[0055] Example 4

[0056] Embodiment 4 of the present invention discloses a method for estimating the health of a converter valve. The converter valve is generally a DC converter valve. Figure 4 As shown, the method comprises the following steps:

[0057] Step S401: Acquire historical monitoring data of a first preset number of converter valves.

[0058] The historical monitoring data is the same as the real-time monitoring data. For details, please refer to the description of Example 2, which will not be repeated here. However, the historical monitoring data is the data collected and stored first. For example, the first preset number is 100.

[0059] Step S402: collecting real-time monitoring data of the converter valve.

[0060] This step is the same as step S201 and will not be described in detail here. The real-time monitoring data is generally the current value.

[0061] It should be understood that, whether it is historical monitoring data or real-time monitoring data, the interval time of collecting adjacent data is the same. The next data after the last data of the acquired historical monitoring data may be exactly the real-time monitoring data.

[0062] Step S403: Use a first preset amount of historical monitoring data and a real-time monitoring data to perform data fitting to generate a short-term health prediction model.

[0063] The data fitting method may be a linear fitting method or other machine learning algorithms.

[0064] Specifically, the short-term health prediction model of an embodiment of the present invention is as follows:

[0065]

[0066] Among them, SS0 is real-time monitoring data, A i is the i-th preset short-term prediction coefficient, which can be determined based on experience, SS i is the i-th historical monitoring data, SS is the predicted short-term monitoring data, and k1 is the first preset number.

[0067] Furthermore, the newly collected real-time monitoring data can be continuously used to verify and correct the short-term health prediction model, so as to correct the short-term health prediction model in real time and realize model iteration to ensure that it can output more accurate values; among which, the verification is mainly to determine whether the data is complete by comparing the total number and the data length set by the data header; the correction is mainly to obtain the missing data by averaging the previous data and the next data.

[0068] Step S404: using the short-term health model to estimate the short-term monitoring data of the converter valve.

[0069] Use the short-term health prediction data model to estimate future short-term monitoring data.

[0070] Step S405: Count the second number of short-term monitoring data that exceeds the range defined by the protection data and the rated data.

[0071] Similarly, when taking statistics, the same type of short-term monitoring data is also counted together with the protection data and rated data.

[0072] Step S406: Calculate a second ratio of the second quantity to the total quantity of all short-term monitoring data.

[0073] This calculation ratio is also calculated in the same category. All short-term monitoring data are all short-term monitoring data within the second preset time range in the future output by the short-term health prediction data model. Generally, the second preset time can be determined by the time span of the real-time monitoring data involved in the estimation, that is, the time interval between the first real-time monitoring data involved in the estimation and the last real-time monitoring data is the second preset time.

[0074] Step S407: estimating a second health level of the converter valve according to the second ratio.

[0075] The second health level can be preset. For example, the second health level can be divided into healthy, slightly unhealthy, and severely unhealthy. Each second health level corresponds to a numerical range. According to the numerical range in which the second ratio is located, the second health level is estimated.

[0076] Specifically, the second health level is estimated using the following calculation formula:

[0077]

[0078] Wherein, J2 is the second health level, D2 is the second ratio, J2=2 corresponds to severe unhealth, J2=1 corresponds to mild unhealth, and J2=0 corresponds to health.

[0079] Embodiment 4 of the present invention can establish a fitting mathematical model through a certain amount of historical monitoring data and real-time monitoring data to estimate data, and use the estimated data to evaluate the health level.

[0080] Example 5

[0081] Embodiment 5 of the present invention discloses a method for estimating the health of a converter valve. The converter valve is generally a DC converter valve. Figure 5 As shown, the method comprises the following steps:

[0082] Step S501: Acquire historical monitoring data of a second preset number of converter valves.

[0083] The historical monitoring data and the real-time monitoring data are the same, and the details can be found in the description of Example 2, which will not be repeated here. However, the historical monitoring data is the data collected and stored first. The second preset number should be greater than the first preset number, for example, the second preset number is 1000.

[0084] Step S502: collecting real-time monitoring data of the converter valve.

[0085] This step is the same as step S201 and will not be described in detail here. The real-time monitoring data is generally the current value.

[0086] It should be understood that, whether it is historical monitoring data or real-time monitoring data, the interval time of collecting adjacent data is the same. The next data after the last data of the acquired historical monitoring data may be exactly the real-time monitoring data.

[0087] Step S503: Use a second preset amount of historical monitoring data and a real-time monitoring data to perform data fitting to generate a long-term health prediction model.

[0088] The data fitting method may be a linear fitting method or other machine learning algorithms.

[0089] Specifically, the long-term health prediction model of an embodiment of the present invention is as follows:

[0090]

[0091] Among them, SS0 is real-time monitoring data, B i is the i-th preset long-term prediction coefficient, determined based on experience, SS i is the i-th historical monitoring data, SL is the predicted long-term monitoring data, and k2 is the second preset number.

[0092] Step S504: using a long-term health prediction model to predict long-term monitoring data of the converter valve.

[0093] Furthermore, the future long-term monitoring data predicted by the long-term health prediction model can be used as iterative prediction data, the earliest historical monitoring data can be deleted, and the iterative prediction data can replace the earliest historical monitoring data. A new prediction can be made through the long-term health prediction model, and so on. The newly predicted long-term monitoring data continuously iterates the earliest historical monitoring data that still exists. After repeating the iteration a preset number of times, a preset number of long-term monitoring data can be generated and saved. The preset number can be determined according to actual needs. For example, the preset number can be 5000, which will generate 5000 long-term monitoring data, i.e., iterative prediction data.

[0094] Step S505: Count the third number of long-term monitoring data that exceeds the range defined by the protection data and the rated data.

[0095] Similarly, when taking statistics, the same type of long-term monitoring data is also counted together with the protection data and rated data.

[0096] Step S506: Calculate a third ratio of the third quantity to the total quantity of all long-term monitoring data.

[0097] This calculation ratio is also calculated in the same category. All long-term monitoring data is a preset number of all long-term monitoring data output by the long-term health prediction data model.

[0098] Step S507: estimating a third health level of the converter valve according to the third ratio.

[0099] The third health level can be preset. For example, the third health level can be divided into healthy, slightly unhealthy, and severely unhealthy. Each third health level corresponds to a numerical range. According to the numerical range in which the third ratio is located, the third health level is estimated.

[0100] Specifically, the third health level is estimated using the following calculation formula:

[0101]

[0102] Wherein, J3 is the third health level of the converter valve, and D3 is the third ratio. J3=2 corresponds to severe unhealthiness, J3=1 corresponds to mild unhealthiness, and J3=0 corresponds to healthiness.

[0103] The numerical ranges of the first health level, the second health level and the third health level mentioned above may be completely the same or different.

[0104] Embodiment 5 of the present invention can estimate data by establishing a fitting mathematical model through a certain amount of historical monitoring data and real-time monitoring data, and replace the earliest historical monitoring data with the estimated data to continuously iterate and estimate the next estimated data. Ultimately, all data are obtained through estimation, and the estimated data is used to evaluate the health level.

[0105] Example 6

[0106] Embodiment 6 of the present invention discloses a computer-readable storage medium, on which computer program instructions are stored; when the computer program instructions are executed by a processor, the method for estimating the health of a converter valve as described in the above embodiment is implemented.

[0107] Example 7

[0108] Embodiment 7 of the present invention discloses a system for estimating the health of a converter valve, including: a computer-readable storage medium as described in the above embodiments.

[0109] The system of Example 7 can sort the data that exceeds the range specified by the protection data and the rated data in the above-mentioned Examples 3 to 5 according to the preset health level of the data, generate a health status view of the converter valves, and can also save the ratios that exceed the specified range as a comprehensive converter valve health level table.

[0110] In summary, the embodiments of the present invention design different judgment logics and use multi-dimensional data to estimate the health of the converter valve, which is highly efficient and helps to timely discover potential abnormalities of the converter valve for prevention.

[0111] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for estimating the health of a converter valve, characterized in that: include: Obtain monitoring data of the converter valve; According to the relationship between the monitoring data, the protection data and the rated data, the health of the converter valve is estimated. The protection data can be obtained by preset, which represents the data when the protection device of the ultra-high voltage direct current transmission is in action. The rated data is the preset normal data. The step of obtaining the monitoring data of the converter valve comprises: Collecting real-time monitoring data of the converter valve; The step of estimating the health of the converter valve according to the magnitude relationship between the monitoring data, the protection data and the rated data comprises: If the real-time monitoring data is closer to the protection data, it is estimated that the health of the converter valve is low; If the real-time monitoring data is closer to the rated data, it is estimated that the health level of the converter valve is high; Alternatively, the step of estimating the health of the converter valve according to the magnitude relationship between the monitoring data, the protection data and the rated data includes: A first quantity of the real-time monitoring data exceeding the range defined by the protection data and the rated data is counted, and the first quantity is counted using the following calculation formula: ; Where D represents the first quantity, S i represents the i-th real-time monitoring data, B i represents the protection data corresponding to the i-th real-time monitoring data, E i represents the rated data corresponding to the i-th real-time monitoring data, the vertical line represents or, and the formula represents S i With B i The difference and E i With S i If one of the differences does not meet the requirements, and the real-time monitoring data exceeds the range of the protection data and the rated data, it will be counted once; A first ratio of the first quantity to the total quantity of all the real-time monitoring data is calculated, and a first health level is estimated using the following calculation formula: , where J1 is the first health level and D1 is the first ratio; estimating a first health level of the converter valve according to the first ratio; The step of obtaining the monitoring data of the converter valve comprises: Acquiring a first preset number of historical monitoring data of the converter valve; Collecting real-time monitoring data of the converter valve; Using a first preset number of the historical monitoring data and one of the real-time monitoring data to perform data fitting to generate a short-term health prediction model; Using the short-term health prediction model to predict short-term monitoring data of the converter valve; The short-term health prediction model is: ； Among them, SS0 is the real-time monitoring data, A i is the i-th preset short-term prediction coefficient, SS i is the i-th historical monitoring data, SS is the predicted short-term monitoring data, and k1 is the first preset number; Alternatively, the step of obtaining the monitoring data of the converter valve includes: Acquiring a second preset amount of historical monitoring data of the converter valve; Collecting real-time monitoring data of the converter valve; Using the second preset number of historical monitoring data and the one real-time monitoring data to perform data fitting to generate a long-term health prediction model; Using the long-term health prediction model to predict long-term monitoring data of the converter valve; The long-term health prediction model is: ； Among them, SS0 is the real-time monitoring data, B i is the i-th preset long-term prediction coefficient, SS i is the i-th historical monitoring data, SL is the predicted long-term monitoring data, and k2 is the second preset number.

2. The method for estimating the health of a converter valve according to claim 1, characterized in that: The step of estimating the health of the converter valve according to the magnitude relationship between the monitoring data, the protection data and the rated data comprises: Counting a second amount of the short-term monitoring data exceeding the range defined by the protection data and the rated data; Calculating a second ratio of the second quantity to the total quantity of all the short-term monitoring data; A second health level of the converter valve is estimated according to the second ratio.

3. The method for estimating the health of a converter valve according to claim 1, characterized in that: The step of estimating the health of the converter valve according to the magnitude relationship between the monitoring data, the protection data and the rated data comprises: Counting a third number of the long-term monitoring data exceeding the range defined by the protection data and the rated data; Calculating a third ratio of the third quantity to the total quantity of all the long-term monitoring data; The health level of the converter valve is estimated according to the third ratio.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, the method for estimating the health of the converter valve according to any one of claims 1 to 3 is implemented.

5. A system for estimating the health of a converter valve, characterized in that: include: The computer readable storage medium of claim 4.

Images