MMC converter valve multi-operation-parameter combined temperature online monitoring method and MMC converter valve multi-operation-parameter combined temperature online monitoring system

Abstract

The invention discloses an MMC converter valve multi-operation parameter combined temperature online monitoring method and system, and the method comprises the steps: obtaining converter valve infrared inspection data, valve cooling system operation data, valve control sub-module monitoring temperature data and operation condition historical data of each valve hall in a converter station, and building a discrete data set; determining an over-temperature alarm fixed value of the temperature of the sub-module component and a relative similar abnormal temperature rise fixed value of the sub-module component; establishing a temperature prediction model according to the discrete data set; monitoring the inspection data, and judging whether temperature abnormity occurs or not through an over-temperature criterion and a temperature prediction model; and generating corresponding over-temperature alarm fault information of the converter valve according to an analysis result, and giving a disposal suggestion. According to the technical scheme, the functions of sub-module power unit temperature monitoring, over-temperature warning and abnormal temperature pre-judgment are achieved through MMC converter valve temperature online monitoring of multi-temperature system monitoring and multi-operation-condition parameter conjoint analysis, and the method has important significance on diversified detection means, enriching detection standards and improving the operation and detection intelligence degree.

Description

Technical Field

[0001] The present invention belongs to the technical field of online monitoring of converter valves, and in particular relates to an online temperature monitoring method and system for combining multiple operating parameters of an MMC converter valve. Background Art

[0002] Flexible DC transmission technology based on modular multilevel converter (MMC) using fully controlled insulated gate bipolar transistor (IGBT) has been applied in a series of DC demonstration projects. Operation and maintenance issues, especially online temperature monitoring of modular power devices of converter valves, have become the focus of attention.

[0003] DL / T 351-2010 "Guidelines for Maintenance of Converter Valves" stipulates the specific implementation methods for three types of inspection work for thyristor converter valves: daily inspection, routine maintenance and special maintenance. There is no reference guide for IGBT converter valves. At present, there are two monitoring methods for IGBT converter valve equipment: one is to measure the surface temperature of the converter valve based on an infrared temperature measurement unit or a handheld infrared temperature measurement instrument by the operation and maintenance personnel. This method cannot deeply detect the temperature of the internal components of the submodule. The measurement accuracy is affected by the measurement distance and measurement angle, and the temperature state relies on manual judgment, with a single standard; the other is to measure the temperature by installing a sensor at the heat sink position of the modular power unit. This method measures the temperature close to the junction temperature of the IGBT device, but the sensor is prone to damage and false alarms. The monitoring function is built into the valve control system. Due to the limitation of the hardware load rate, only the five power units with the highest temperature in each valve tower can be monitored. In addition, a method has been proposed to indirectly infer the temperature of the converter valve by monitoring the temperature of the cooling medium at the inlet and outlet of the cooling system. This method has not been implemented in engineering so far, and its effectiveness has not been verified.

[0004] In summary, with the development of high-voltage and large-capacity IGBT converter valves, online temperature monitoring and timely warning of converter valves are particularly important. However, the existing online temperature monitoring methods of converter valves have problems such as single measurement means, simple judgment standards, and low intelligent operation and inspection level. Therefore, the online temperature monitoring of MMC converter valves through multi-temperature system monitoring and joint analysis of multiple operating condition parameters realizes sub-module power unit temperature monitoring, over-temperature alarm, and abnormal temperature prediction functions, which is of great significance to diversify detection methods, enrich detection standards, and improve the intelligent level of operation and inspection. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for online temperature monitoring of an MMC converter valve with multiple operating parameters. The online temperature monitoring of the MMC converter valve with multiple temperature system monitoring and joint analysis of multiple operating condition parameters realizes submodule power unit temperature monitoring, overtemperature alarm, and abnormal temperature prediction functions, which is of great significance for diversifying detection methods, enriching detection standards, and improving the intelligent level of operation and inspection.

[0006] In order to achieve the above object, the solution of the present invention is:

[0007] A temperature online monitoring method for combining multiple operating parameters of an MMC converter valve, comprising:

[0008] Acquire the infrared inspection data of the converter valves of each valve hall in the converter station, the operation data of the valve cooling system, the temperature data monitored by the valve control submodule and the historical data of the operation condition, establish a discrete data set, and establish a temperature prediction model based on the discrete data set;

[0009] Determine the over-temperature alarm setting value of the sub-module component temperature and the abnormal temperature rise setting value of the sub-module component relative to the same type, and construct the over-temperature judgment criterion;

[0010] Based on real-time inspection data, the temperature prediction model is used to calculate the temperature prediction interval, and whether the temperature of the submodule is normal is judged according to whether the actual temperature of the converter valve submodule falls into the temperature prediction interval; the over-temperature criterion is used to judge whether the converter valve has absolute temperature over-limit and relative temperature over-limit faults.

[0011] The infrared inspection data of the converter valves in each valve hall in the converter station, the operation data of the valve cooling system, the temperature monitoring data of the valve control submodule and the historical data of the operating conditions are obtained to establish a discrete data set, including the end time of two consecutive infrared scheduled inspections as the start and end intervals of the data set, and the operation status of the infrared inspection system of the converter valves in each valve hall in the converter station within the time period, the operation status of the valve cooling system, the temperature monitoring and operating condition historical information of the valve control system submodule, so as to establish a discrete data set of characteristic quantities.

[0012] Establishing a temperature prediction model according to the discrete data set includes:

[0013] Determine a data set {X1, X2, X3, ...} of input feature parameters and a data set {Y1, Y2, ...} of output feature parameters, perform correlation analysis on the data sets, and use the input feature parameters and output feature parameters that are higher than a set threshold as input data;

[0014] Preprocessing and standardizing the input data;

[0015] The processed input data is divided into training set and test set, and convolutional neural network is used for model prediction;

[0016] The training model uses MSE, RMSE, MAE and R 2 Four evaluation indicators are used to evaluate the model until it is completed as a temperature prediction model; MSE represents mean square error, RMSE represents root mean square error, MAE represents mean absolute error, R 2 represents the coefficient of determination.

[0017] Judging whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval includes:

[0018] After removing abnormal values ​​from the real-time inspection data, the submodule temperature online monitoring system, valve cooling system operating status and operating condition data sets are obtained, and the temperature prediction range of this inspection task is calculated using the temperature prediction model;

[0019] It is determined whether the actual temperature of the converter valve submodule falls within the temperature prediction interval. If it does not fall within the interval, it is determined that the actual temperature of the converter valve submodule exceeds the predicted temperature rise; otherwise, it is determined that the temperature of the converter valve submodule is normal.

[0020] Determine the submodule component temperature over-temperature alarm setting value and the submodule component relative similar abnormal temperature rise setting value, and construct an over-temperature criterion, including the over-temperature criterion including an absolute temperature over-limit criterion and a relative temperature over-limit criterion;

[0021] The absolute temperature over-limit criterion includes: any point temperature value ≥ M1, which is determined as the absolute temperature over-limit level III alarm of the converter valve; M1>any point temperature ≥ M2, which is determined as the absolute temperature over-limit level II alarm of the converter valve; wherein M1 is the first set value of the submodule component temperature over-temperature alarm, M2 is the second set value of the submodule component temperature over-temperature alarm, and M1>M2;

[0022] The relative temperature over-limit criterion includes: the temperature difference between submodules ≥ N1, which is judged as the relative temperature over-limit level II alarm of the flow valve submodule; the relative temperature difference of the copper busbar joint of the submodule ≥ N2, which is judged as the relative temperature over-limit level II alarm of the copper busbar joint of the converter valve; wherein N1 is the first fixed value of the abnormal temperature rise of the submodule component relative to the same type, and N2 is the second fixed value of the abnormal temperature rise of the submodule component relative to the same type.

[0023] Use the over-temperature criterion to determine whether the converter valve has an absolute temperature over-limit, including:

[0024] Get patrol temperature values ​​in real time;

[0025] The inspection temperature values ​​are updated in real time through polling and judged one by one.

[0026] Based on the real-time inspection data, before making a judgment, the real-time inspection data is cleaned, including:

[0027] Analyze all data points of each camera during one inspection and count the same temperature points. When the number of the same temperature points exceeds half of the total temperature data, it is determined that the PTZ may be faulty.

[0028] For cameras that are judged to be possibly faulty, the current inspection data is used as the starting point, and three consecutive inspection data are calculated forward to determine whether the number of identical temperature points each time exceeds half of the total temperature data; at the same time, the three inspection data are compared according to the point positions. If they are all the same, it is determined that the infrared pan / tilt head is faulty.

[0029] Whether the temperature of the submodule is normal is determined based on whether the actual temperature of the converter valve submodule falls within the temperature prediction interval, and / or, an over-temperature criterion is used to determine whether the converter valve has an absolute temperature over-limit and a relative temperature over-limit fault. Thereafter, corresponding alarm fault information is generated based on the judgment result.

[0030] An online temperature monitoring system for MMC converter valves with multiple operating parameters, comprising:

[0031] The temperature prediction model building module is configured to obtain infrared inspection data of converter valves in each valve hall in the converter station, valve cooling system operation data, valve control submodule monitoring temperature data and operation condition history data, establish a discrete data set, and establish a temperature prediction model according to the discrete data set;

[0032] An over-temperature judgment criterion building module is configured to determine an over-temperature alarm setting value of a sub-module component temperature and a relative abnormal temperature rise setting value of a sub-module component of the same type, and build an over-temperature judgment criterion;

[0033] A submodule temperature fault judgment module is configured to calculate a temperature prediction interval based on real-time inspection data using a temperature prediction model, and judge whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval; and

[0034] The absolute temperature over-limit and relative temperature over-limit fault judgment module is configured to judge whether the converter valve has an absolute temperature over-limit and relative temperature over-limit fault based on real-time inspection data and using over-temperature judgment criteria.

[0035] The temperature prediction model construction module obtains the infrared inspection data of the converter valves in each valve hall in the converter station, the operation data of the valve cooling system, the temperature monitoring data of the valve control submodule and the historical data of the operating conditions, and establishes a discrete data set, including the end time of two consecutive infrared scheduled inspections as the start and end intervals of the data set, and obtains the operation status of the infrared inspection system of the converter valves in each valve hall in the converter station within the time period, the operation status of the valve cooling system, the temperature monitoring and operating condition historical information of the valve control system submodule, thereby establishing a discrete data set of characteristic quantities.

[0036] The temperature prediction model building module establishes a temperature prediction model according to the discrete data set, including:

[0037] Determine a data set {X1, X2, X3, ...} of input feature parameters and a data set {Y1, Y2, ...} of output feature parameters, perform correlation analysis on the data sets, and use the input feature parameters and output feature parameters that are higher than a set threshold as input data;

[0038] Preprocessing and standardizing the input data;

[0039] The processed input data is divided into training set and test set, and convolutional neural network is used for model prediction;

[0040] The training model uses MSE, RMSE, MAE and R 2 Four evaluation indicators are used to evaluate the model until it is completed as a temperature prediction model; MSE represents mean square error, RMSE represents root mean square error, MAE represents mean absolute error, R 2 represents the coefficient of determination.

[0041] The submodule temperature fault judgment module judges whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval, including:

[0042] After removing abnormal values ​​from the real-time inspection data, the submodule temperature online monitoring system, valve cooling system operating status and operating condition data sets are obtained, and the temperature prediction range of this inspection task is calculated using the temperature prediction model;

[0043] It is determined whether the actual temperature of the converter valve submodule falls within the temperature prediction interval. If it does not fall within the interval, it is determined that the actual temperature of the converter valve submodule exceeds the predicted temperature rise; otherwise, it is determined that the temperature of the converter valve submodule is normal.

[0044] The over-temperature criterion building module determines the over-temperature alarm setting value of the sub-module component temperature and the relative abnormal temperature rise setting value of the sub-module component of the same type, and builds the over-temperature criterion, including the over-temperature criterion including the absolute temperature over-limit criterion and the relative temperature over-limit criterion;

[0045] The absolute temperature over-limit criterion includes: any point temperature value ≥ M1, which is determined as the absolute temperature over-limit level III alarm of the converter valve; M1>any point temperature ≥ M2, which is determined as the absolute temperature over-limit level II alarm of the converter valve; wherein M1 is the first set value of the submodule component temperature over-temperature alarm, M2 is the second set value of the submodule component temperature over-temperature alarm, and M1>M2;

[0046] The relative temperature over-limit criterion includes: the temperature difference between submodules ≥ N1, which is judged as the relative temperature over-limit level II alarm of the flow valve submodule; the relative temperature difference of the copper busbar joint of the submodule ≥ N2, which is judged as the relative temperature over-limit level II alarm of the copper busbar joint of the converter valve; wherein N1 is the first fixed value of the abnormal temperature rise of the submodule component relative to the same type, and N2 is the second fixed value of the abnormal temperature rise of the submodule component relative to the same type.

[0047] The above-mentioned absolute temperature over-limit and relative temperature over-limit fault judgment module uses the over-temperature judgment criteria to judge whether the converter valve has an absolute temperature over-limit, including:

[0048] Get patrol temperature values ​​in real time;

[0049] The inspection temperature values ​​are updated in real time through polling and judged one by one.

[0050] The system further comprises,

[0051] The data cleaning module is configured to clean the real-time inspection data before the submodule temperature fault judgment module, the absolute temperature over-limit and the relative temperature over-limit fault judgment module make judgments based on the real-time inspection data, including:

[0052] Analyze all data points of each camera during one inspection and count the same temperature points. When the number of the same temperature points exceeds half of the total temperature data, it is determined that the PTZ may be faulty.

[0053] For cameras that are judged to be possibly faulty, the current inspection data is used as the starting point, and three consecutive inspection data are calculated forward to determine whether the number of identical temperature points each time exceeds half of the total temperature data; at the same time, the three inspection data are compared according to the point positions. If they are all the same, it is determined that the infrared pan / tilt head is faulty.

[0054] The system further comprises,

[0055] The fault warning module is configured to generate corresponding alarm fault information according to the judgment results after the submodule temperature fault judgment module makes a judgment and / or after the absolute temperature over-limit and relative temperature over-limit fault judgment modules make a judgment.

[0056] A computer device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor; when the processor executes the computer program, the steps of the above-mentioned method for online temperature monitoring of multiple operating parameters of an MMC converter valve are implemented.

[0057] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the steps of the above-mentioned method for online temperature monitoring of multiple operating parameters of an MMC converter valve are implemented.

[0058] After adopting the above scheme, the beneficial effects of the present invention are as follows:

[0059] The present invention first obtains the characteristic quantities of the infrared timing inspection system, valve cooling system, valve control system and operating conditions of the converter valves in each valve hall of the converter station to construct a discrete data set; then, according to the characteristics of the infrared inspection data, the over-temperature alarm setting value of the sub-module component temperature and the abnormal temperature rise setting value of the sub-module component relative to the same type are determined; then, according to the discrete set of characteristic quantities, a sub-module temperature model based on a convolutional neural network is established to realize the temperature interval [T max ,T min ] prediction; monitor infrared timing tasks, analyze temperature data based on the above-mentioned fixed values ​​and temperature ranges, and determine whether there are abnormal temperature measurement points; finally, judge the operating status of the converter valve according to the above three types of criteria, distinguish the severity, generate corresponding over-temperature alarms, and formulate corresponding maintenance strategies to achieve intelligent decision-making.

[0060] (1) Diversified detection methods

[0061] At present, the temperature monitoring of the converter valve hall relies on the valve hall infrared camera regular inspection task and the converter valve monitoring background to monitor the maximum five submodules and the minimum five submodules of each converter valve temperature. The two types of detection systems independently monitor the temperature changes in the valve hall without data interaction. The present invention proposes to establish a convolutional neural network submodule temperature prediction model based on the valve cooling-valve control-operation condition multi-parameter combination based on the discrete set of infrared inspection data, valve cooling operation data, valve control submodule temperature monitoring data and historical operating conditions to achieve the prediction of the temperature interval [Tmax, Tmin], and jointly analyze the multiple operation and maintenance parameters of the valve hall system, enhance the data interaction capability, and improve the possibility of fault analysis and diagnosis.

[0062] (2) Enriching testing standards

[0063] The existing over-temperature alarm is mainly used in infrared scheduled inspections. After setting the fixed value of the over-temperature alarm, it only alarms the measurement points that exceed the fixed value. The single criterion cannot meet the operation and maintenance needs. The present invention sets the over-temperature fault criterion from three aspects: the absolute value, relative value and predicted value of temperature. The over-temperature fault is divided into levels according to the severity of the over-temperature fault. The corresponding disposal suggestions include maintenance at the appropriate time, application for immediate maintenance, and emergency stop maintenance, which provides a basis for operation and maintenance.

[0064] (3) Improving intelligent operation and inspection

[0065] The current over-temperature judgment criteria are limited to the cross-boundary comparison of a single fixed value, and the intelligent means are insufficient. The convolutional neural network sub-module temperature model based on the valve cooling-valve control-operating condition multi-parameter combination adds a prediction function on the basis of the fixed value judgment criteria, which meets the two aspects of operation and inspection requirements of the online monitoring and prediction functions of the converter valve temperature, and realizes the sub-module power unit temperature monitoring, over-temperature alarm, and abnormal temperature prediction functions, which is of great significance to operation and inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flow chart of a method for online temperature monitoring of multiple operating parameters of an MMC converter valve in a preferred embodiment of the present invention;

[0067] Figure 2 is a calculation flow chart of over-temperature fault judgment in a preferred embodiment of the present invention;

[0068] Figure 3 It is a schematic diagram of an online temperature monitoring system and its working mode in combination with multiple operating parameters of an MMC converter valve in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0069] 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 only part of the embodiments of the present invention. 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.

[0070] like Figure 1 As shown, the present invention provides a method for online temperature monitoring of multiple operating parameters of an MMC converter valve, comprising:

[0071] Acquire the infrared inspection data of the converter valves of each valve hall in the converter station, the operation data of the valve cooling system, the temperature data monitored by the valve control submodule and the historical data of the operation condition, establish a discrete data set, and establish a temperature prediction model based on the discrete data set;

[0072] Determine the over-temperature alarm setting value of the sub-module component temperature and the abnormal temperature rise setting value of the sub-module component relative to the same type, and construct the over-temperature judgment criterion;

[0073] Based on real-time inspection data, the temperature prediction model is used to calculate the temperature prediction interval, and whether the temperature of the submodule is normal is judged according to whether the actual temperature of the converter valve submodule falls into the temperature prediction interval; the over-temperature criterion is used to judge whether the converter valve has absolute temperature over-limit and relative temperature over-limit faults.

[0074] The present invention also provides an online temperature monitoring system for MMC converter valves with multiple operating parameters, comprising:

[0075] The temperature prediction model building module is configured to obtain infrared inspection data of converter valves in each valve hall in the converter station, valve cooling system operation data, valve control submodule monitoring temperature data and operation condition history data, establish a discrete data set, and establish a temperature prediction model according to the discrete data set;

[0076] An over-temperature judgment criterion building module is configured to determine an over-temperature alarm setting value of a sub-module component temperature and a relative abnormal temperature rise setting value of a sub-module component of the same type, and build an over-temperature judgment criterion;

[0077] A submodule temperature fault judgment module is configured to calculate a temperature prediction interval based on real-time inspection data using a temperature prediction model, and judge whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval; and

[0078] The absolute temperature over-limit and relative temperature over-limit fault judgment module is configured to judge whether the converter valve has an absolute temperature over-limit and relative temperature over-limit fault based on real-time inspection data and using over-temperature judgment criteria.

[0079] The technical solutions and beneficial effects of the present invention are described in detail below through specific embodiments.

[0080] The embodiment of the present invention proposes a temperature online monitoring method for MMC converter valves in combination with multiple operating parameters. The monitoring part determines the absolute temperature crossing alarm when the temperature measurement value exceeds the maximum allowable temperature value and the relative temperature crossing alarm when a certain component has an abnormal temperature rise relative to other components during the same inspection based on DL / T 351-2010 "Guidelines for the Maintenance of Converter Valves" and the temperature characteristics of the MMC converter valve itself; the absolute temperature crossing alarm is divided into the absolute temperature crossing alarm of the converter valve at level III and the absolute temperature crossing alarm of the converter valve at level II according to the set maximum allowable temperature value; the relative temperature crossing alarm is divided into the relative temperature crossing alarm of the converter valve submodule at level II and the relative temperature crossing alarm of the converter valve copper bar joint at level II according to the different components. The prediction part establishes a convolutional neural network temperature prediction model based on the infrared inspection data set, the valve cooling system operation data set, the valve control submodule monitoring temperature data set and the historical operating condition data set to realize the temperature prediction of multiple operating parameters and predict temperature abnormalities.

[0081] In the first preferred embodiment of the present invention, infrared inspection data of converter valves in each valve hall in the converter station, operation data of the valve cooling system, temperature data monitored by the valve control submodule and historical data of the operation condition are obtained to establish a discrete data set, and a temperature prediction model is established based on the discrete data set;

[0082] In this embodiment, when establishing a discrete data set, the end time of two consecutive infrared timing inspections is used as the start and end interval of the data set, that is, the end time of the previous (n-1th) infrared timing inspection task is used as the start time, recorded as t start, the end time of this (nth) infrared scheduled inspection task is taken as the end time, recorded as t end ; Obtain the operating status of the infrared inspection system of the converter valves in each valve hall in the converter station during this time period, the operating status of the valve cooling system, the monitoring temperature of the valve control system submodule and the historical information of the operating conditions, so as to establish a discrete data set of characteristic quantities.

[0083] Among them, the characteristic quantities of the four types of data, namely, infrared inspection data, valve cooling system operation data, valve control submodule monitoring temperature data and operation condition data, include: start , t end ] interval,

[0084] Infrared timing inspection system: The maximum temperature of the m preset positions of the infrared nth timing inspection of the converter valve, T infrared ={preset maximum temperature};

[0085] Valve cooling system: water-cooled inlet valve temperature data set T cooling_in ={}, water cooling valve temperature data set T cooling_out ={}, cooling water flow data set C cooling ={};

[0086] Valve control submodule monitoring temperature: Maximum five submodule temperature data sets T SM_MAX ={[T MAX_NO1 ],[T MAX_NO2 ],[T MAX_NO3 ],[T MAX_NO4 ],[T MAX_NO5 ]}, where T MAX_NO1 is the highest submodule temperature among the five modules, T MAX_NO2 is the temperature of the submodule with the second highest temperature among the five modules, and so on; the minimum five submodule temperature data set T SM_MIN ={[T MIN_NO1 ],[T MIN_NO2 ],[T MIN_NO3 ],[T MIN_NO4 ],[T MIN_NO5 ]}, where T MIN_NO1 is the temperature of the lowest submodule among the five modules, T MIN_NO2 is the temperature of the submodule with the second lowest temperature among the five modules, and so on;

[0087] Operating conditions: converter valve active power data set P = {}, converter valve hall temperature data set T hall ={}.

[0088] In the first embodiment, the flow chart for constructing the temperature prediction model includes the following steps:

[0089] Determine the data set {X1, X2, X3, ...} of the input feature parameters and the data set {Y1, Y2, ...} of the output feature parameters, perform correlation analysis, and select the feature parameters as input;

[0090] The selected input features include:

[0091] Valve inlet temperature data set of valve cooling system, T cooling_in ={};

[0092] Valve outlet temperature data set of valve cooling system, T cooling_out ={};

[0093] Cooling water flow data set of valve cooling system, C cooling ={};

[0094] Submodule monitoring maximum submodule temperature data set, T MAX_NO1 ={};

[0095] Submodule monitoring minimum submodule temperature data set, T MIN_NO1 ={};

[0096] Converter valve active power data set, P = {};

[0097] Converter valve hall temperature data set, T hall ={}.

[0098] The selected output features include:

[0099] The maximum and minimum values ​​of the set of the highest temperatures of all preset positions of the converter valve infrared nth timed inspection, T infrared ={T MAX ,T MIN}

[0100] Input data preprocessing, data default value and abnormal data analysis, and standardization. Since the sampling frequency of the above eigenvalues ​​is consistent, the length of the constructed data sequence is equal, so there is no need for default value processing; for abnormal values, the 3σ criterion is used to eliminate the error data set; the remaining data set is standardized.

[0101] The processed input feature quantities and output feature quantities are divided into training sets and test sets, and convolutional neural networks (CNN) are used for model prediction.

[0102] The training model uses MSE, RMSE, MAE and R 2 Four evaluation indicators are used to evaluate until the model is completed.

[0103] In the second preferred embodiment of the present invention, the over-temperature alarm setting value of the submodule component temperature and the relative abnormal temperature rise setting value of the submodule component of the same type are determined to construct an over-temperature judgment criterion;

[0104] In this embodiment, based on DL / T 351-2010 "Guidelines for Maintenance of Converter Valves" and the temperature characteristics of the MMC converter valve itself, an absolute temperature out-of-bounds alarm is determined when the temperature measurement value exceeds the maximum allowable temperature value, and a relative temperature out-of-bounds alarm is determined when a certain component has an abnormal temperature rise relative to other components during the same inspection.

[0105] For the over-temperature alarm setting, two over-temperature criteria are set:

[0106] ① The absolute temperature of the converter valve exceeds the limit and the alarm is level III. The set value is M1: the temperature value of any point ≥ M1

[0107] ② The absolute temperature of the converter valve exceeds the limit and the alarm is level II. The set value is M2: M1>Any point temperature ≥ M2

[0108] The infrared inspection objects of the MMC converter valve sub-components are the sub-module surface temperature and the copper busbar joint temperature. For the same inspection data, the over-temperature judgment criteria are set respectively:

[0109] ① The relative temperature of the converter valve submodule exceeds the limit level II alarm: the submodule IGBT and capacitor have abnormal heating, relative to the same type of abnormal temperature rise set value N1, the temperature difference between modules ≥ N1

[0110] ② The relative temperature of the copper busbar joint of the converter valve exceeds the limit. Level II alarm: The copper busbar joint of the submodule is abnormally heated, and the relative temperature difference is ≥ N2 relative to the same abnormal temperature rise value.

[0111] In the third preferred embodiment of the present invention, based on real-time inspection data, a temperature prediction model is used to calculate the temperature prediction interval, and whether the submodule temperature is normal is determined based on whether the actual temperature of the converter valve submodule falls within the temperature prediction interval; and an over-temperature criterion is used to determine whether the converter valve has an absolute temperature over-limit and a relative temperature over-limit fault. Figure 2 As shown, the following steps are included:

[0112] Monitor the scheduled task status of the infrared temperature measurement system of the converter valves in each valve hall in the converter station in real time, monitor the start of the scheduled inspection task, obtain the inspection temperature value in real time, and judge whether the surface temperature of the converter valve is abnormally high based on the measured temperature of the submodule. If so, issue an "absolute temperature of the converter valve exceeds the limit" alarm;

[0113] When the scheduled inspection task is detected to have started, the next step is to determine the task execution status; when the scheduled inspection task has not started, the previous step is returned;

[0114] Monitor the execution of this scheduled task to determine which type of fault determination procedure to enter. If this task is being executed, determine whether there is an absolute temperature out-of-bounds fault; if this task has ended, enter the relative temperature out-of-bounds fault determination process;

[0115] The above-mentioned determination of whether there is an absolute temperature out-of-bounds fault is implemented as follows:

[0116] Obtain inspection temperature values ​​in real time;

[0117] Update the inspection data in real time through polling and determine the temperature one by one. According to the relationship between the temperature value and the absolute temperature crossing threshold, if the absolute temperature crosses the limit, a corresponding alarm is issued. If not, continue to obtain the inspection temperature value in real time and re-determine.

[0118] The above-mentioned judgment process of entering the relative temperature out-of-bounds fault includes obtaining all the data of this inspection, performing data cleaning, and eliminating invalid inspection data caused by probe failure. The infrared camera probe of the infrared inspection system may have a probe pan / tilt failure or a probe pan / tilt jam during the inspection. At this time, the infrared probe has been measuring the same pre-set point and the inspection task has not been completed. However, the system still corresponds the sent inspection data one by one according to the preset points, causing data problems. For this type of data, the elimination method is as follows:

[0119] For each inspection data, analyze all data points of each camera inspection and count the same temperature points. When the number of the same temperature points does not exceed half of the total temperature data, it is determined that the PTZ is not faulty and proceed to the next step; when the number of the same temperature points exceeds half of the total temperature data, it is determined that the PTZ may be faulty; at this time, continue to make the following judgments:

[0120] For the infrared camera that may be stuck, the current inspection data is used as the starting point, and three consecutive inspection data are calculated to determine whether the number of the same temperature points each time exceeds half of the total temperature data; at the same time, the three inspection data are compared according to the point positions. If they are the same, it is determined that the infrared pan-tilt head is faulty.

[0121] According to the size relationship of the temperature values ​​of similar components in this inspection, it is determined whether there is a relative temperature out-of-bounds fault. If so, a corresponding alarm is issued. If not, proceed to the next step.

[0122] For the temperature data set of this inspection, outliers are removed, and the sub-module temperature online monitoring system, valve cooling system operating status and operating condition data set are obtained. The temperature prediction range of this inspection task is calculated through the sub-module temperature prediction model based on the multi-parameter combination of valve cooling, valve control and operating condition;

[0123] It is determined whether the actual temperature falls within the predicted interval. If it does not fall within the interval, an alarm of "the actual temperature of the converter valve submodule exceeds the predicted temperature rise" is issued. Otherwise, it is determined that the submodule temperature is normal.

[0124] In the fourth preferred embodiment of the present invention, after judging whether the temperature of the submodule is normal, and / or judging whether the converter valve has an absolute temperature over-limit and a relative temperature over-limit fault, corresponding alarm fault information is generated according to the judgment result and handling suggestions are given.

[0125] In the embodiment of the present invention, the severity of the over-temperature fault is divided into several levels, namely, slight, severe, and emergency, by comparing the actual temperature measured by the converter valve infrared inspection system with different temperature settings of the converter valve.

[0126] The severity definitions are shown in Table 2.

[0127] Table 2 Definition of severity

[0128]

[0129] In the embodiment of the present invention, the converter valve overtemperature alarm information includes the overtemperature valve tower number, the overtemperature submodule number or the location of the copper bar joint, the type and severity of the overtemperature alarm; the overtemperature fault is divided into three levels according to the severity, slight (level I), severe (level II), and fault (level III). The disposal suggestions include maintenance at the appropriate time, application for immediate maintenance, and emergency stop maintenance.

[0130] An embodiment of the present invention further proposes an MMC converter valve multi-operation parameter combined temperature online monitoring system, comprising a converter valve infrared patrol monitoring device, a valve cooling monitoring device, a converter valve monitoring device and an intelligent monitoring platform, wherein the converter valve infrared patrol device receives the start signal and patrol data of one or more sets of infrared cameras in the valve hall for scheduled patrol tasks, the valve cooling monitoring device receives one or more sets of valve cooling equipment belonging to the inlet valve temperature, outlet valve temperature, and coolant flow information, the converter valve monitoring device receives the maximum or minimum submodule temperature of one or more converter valves in a valve hall, as well as the converter valve hall temperature and power information, and the intelligent monitoring platform receives the infrared patrol data of one or more converter valve infrared patrol devices, the data of one or more valve cooling monitoring devices, and the temperature and operating condition data of one or more converter valve monitoring devices, to form discrete data sets of their respective characteristic quantities.

[0131] The intelligent monitoring platform is used to perform temperature prediction based on infrared inspection data, valve cooling system operation data, sub-module monitoring temperature data and historical operating condition data sets; the intelligent monitoring platform also determines whether temperature abnormalities occur based on over-temperature judgment criteria and temperature prediction models, generates corresponding converter valve over-temperature alarm fault information based on the analysis results, and provides disposal suggestions.

[0132] In another preferred embodiment of the present invention, Figure 3 As shown, a single valve hall of a converter station is taken as an example. The temperature online monitoring system for multiple operating parameters of the MMC converter valve in each valve hall includes one converter valve infrared patrol control device, one converter valve monitoring device, and one valve cooling monitoring device. The valve halls of the entire station share one intelligent monitoring platform. The converter valve monitoring device receives the sub-module temperature information and operating condition information of the 12 converter valve towers in a valve hall; the valve cooling monitoring device receives the operation of two sets of valve cooling equipment; the intelligent monitoring platform receives the data sets of the converter valve infrared patrol control device, the converter valve monitoring device, and the valve cooling monitoring device.

[0133] In the embodiment of the present invention, the valve hall infrared inspection camera sends the start and end time of the scheduled inspection task and the inspection data to the intelligent monitoring platform through the valve hall infrared inspection detection device in the form of classified data sets; the submodule temperature of 12 sets of valve towers of the converter valve and the valve hall working condition data are sent to the intelligent monitoring platform through the valve hall converter valve monitoring device in the form of classified data sets; the inlet valve temperature, outlet valve temperature and cooling water flow of 2 sets of valve cooling devices in the valve cooling system are sent to the intelligent monitoring platform through the valve cooling detection device in the form of data sets; the intelligent monitoring platform performs online monitoring of the valve hall temperature using the temperature online monitoring method combined with multiple operating parameters of the MMC converter valve; generates converter valve overtemperature fault alarm information and gives disposal suggestions. Determining the severity of the fault, generating alarm information, and giving disposal suggestions are all completed in the intelligent monitoring platform.

[0134] An embodiment of the present invention also provides another MMC converter valve multiple operating parameters combined temperature online monitoring device, including a processor and a memory configured to store a computer program that can be run on the processor; wherein, when the processor is configured to run the computer program, it executes the method steps in the aforementioned embodiment.

[0135] In practical applications, the processor includes a field programmable gate array (FPGA), and the processor may be a central processing unit (CPU) or a digital signal processor (DSP). It is understandable that for different devices, the electronic device used to implement the function of the processor may be other, and the embodiment of the present invention does not specifically limit it.

[0136] The above-mentioned memory can be a volatile memory (volatile memory), such as a random access memory (RAM); or a non-volatile memory (non-volatile memory), such as a read-only memory (ROM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD); or a combination of the above-mentioned types of memory, and provide instructions and data to the processor.

[0137] In an exemplary embodiment, an embodiment of the present invention further provides a computer-readable storage medium for storing a computer program.

[0138] Optionally, the computer-readable storage medium can be applied to any one of the methods in the embodiments of the present invention, and the computer program enables the computer to execute the corresponding processes implemented by the processor in each method in the embodiments of the present invention. For the sake of brevity, they are not described here.

[0139] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0140] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for online temperature monitoring of multiple operating parameters of an MMC converter valve, characterized in that: include, Acquire the infrared inspection data of the converter valves of each valve hall in the converter station, the operation data of the valve cooling system, the temperature data monitored by the valve control submodule and the historical data of the operation condition, establish a discrete data set, and establish a temperature prediction model based on the discrete data set; Determine the over-temperature alarm setting value of the sub-module component temperature and the abnormal temperature rise setting value of the sub-module component relative to the same type, and construct the over-temperature judgment criterion; Based on real-time inspection data, the temperature prediction model is used to calculate the temperature prediction interval, and whether the temperature of the submodule is normal is judged according to whether the actual temperature of the converter valve submodule falls into the temperature prediction interval; the over-temperature criterion is used to judge whether the converter valve has absolute temperature over-limit and relative temperature over-limit faults.

2. The method according to claim 1, characterized in that: The infrared inspection data of the converter valves in each valve hall in the converter station, the operation data of the valve cooling system, the temperature monitoring data of the valve control submodule and the historical data of the operating conditions are obtained to establish a discrete data set, including the end time of two consecutive infrared scheduled inspections as the start and end intervals of the data set, and the operation status of the infrared inspection system of the converter valves in each valve hall in the converter station within the time period, the operation status of the valve cooling system, the temperature monitoring and operating condition historical information of the valve control system submodule, so as to establish a discrete data set of characteristic quantities.

3. The method according to claim 1, characterized in that: Establishing a temperature prediction model according to the discrete data set includes: Determine a data set {X1, X2, X3, ...} of input feature parameters and a data set {Y1, Y2, ...} of output feature parameters, perform correlation analysis on the data sets, and use the input feature parameters and output feature parameters that are higher than a set threshold as input data; Preprocessing and standardizing the input data; The processed input data is divided into training set and test set, and convolutional neural network is used for model prediction; The training model uses MSE, RMSE, MAE and R 2 Four evaluation indicators are used to evaluate the model until it is completed as a temperature prediction model; MSE represents mean square error, RMSE represents root mean square error, MAE represents mean absolute error, R 2 represents the coefficient of determination.

4. The method according to claim 1, characterized in that: Judging whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval includes: After removing abnormal values ​​from the real-time inspection data, the submodule temperature online monitoring system, valve cooling system operating status and operating condition data sets are obtained, and the temperature prediction range of this inspection task is calculated using the temperature prediction model; It is determined whether the actual temperature of the converter valve submodule falls within the temperature prediction interval. If it does not fall within the interval, it is determined that the actual temperature of the converter valve submodule exceeds the predicted temperature rise; otherwise, it is determined that the temperature of the converter valve submodule is normal.

5. The method according to claim 1, characterized in that: Determine the submodule component temperature over-temperature alarm setting value and the submodule component relative similar abnormal temperature rise setting value, and construct an over-temperature criterion, including the over-temperature criterion including an absolute temperature over-limit criterion and a relative temperature over-limit criterion; The absolute temperature over-limit criterion includes: any point temperature value ≥ M1, which is determined as the absolute temperature over-limit level III alarm of the converter valve; M1>any point temperature ≥ M2, which is determined as the absolute temperature over-limit level II alarm of the converter valve; wherein M1 is the first set value of the submodule component temperature over-temperature alarm, M2 is the second set value of the submodule component temperature over-temperature alarm, and M1>M2; The relative temperature over-limit criterion includes: the temperature difference between submodules ≥ N1, which is judged as the relative temperature over-limit level II alarm of the flow valve submodule; the relative temperature difference of the copper busbar joint of the submodule ≥ N2, which is judged as the relative temperature over-limit level II alarm of the copper busbar joint of the converter valve; wherein N1 is the first fixed value of the abnormal temperature rise of the submodule component relative to the same type, and N2 is the second fixed value of the abnormal temperature rise of the submodule component relative to the same type.

6. The method according to claim 1, characterized in that: Use the over-temperature criterion to determine whether the converter valve has an absolute temperature over-limit, including: Get patrol temperature values ​​in real time; The inspection temperature values ​​are updated in real time through polling and judged one by one.

7. The method according to claim 1, characterized in that: Based on the real-time inspection data, before making a judgment, the real-time inspection data is cleaned. include, Analyze all data points of each camera during one inspection and count the same temperature points. When the number of the same temperature points exceeds half of the total temperature data, it is determined that the PTZ may be faulty. For cameras that are judged to be possibly faulty, the current inspection data is used as the starting point, and three consecutive inspection data are calculated forward to determine whether the number of identical temperature points each time exceeds half of the total temperature data; at the same time, the three inspection data are compared according to the point positions. If they are all the same, it is determined that the infrared pan / tilt head is faulty.

8. The method according to claim 1, characterized in that: Whether the temperature of the submodule is normal is determined based on whether the actual temperature of the converter valve submodule falls within the temperature prediction interval, and / or, an over-temperature criterion is used to determine whether the converter valve has an absolute temperature over-limit and a relative temperature over-limit fault. Thereafter, corresponding alarm fault information is generated based on the judgment result.

9. An online temperature monitoring system for MMC converter valves with multiple operating parameters, characterized in that: include, The temperature prediction model building module is configured to obtain infrared inspection data of converter valves in each valve hall in the converter station, valve cooling system operation data, valve control submodule monitoring temperature data and operation condition history data, establish a discrete data set, and establish a temperature prediction model according to the discrete data set; An over-temperature judgment criterion building module is configured to determine an over-temperature alarm setting value of a sub-module component temperature and a relative abnormal temperature rise setting value of a sub-module component of the same type, and build an over-temperature judgment criterion; The submodule temperature fault judgment module is configured to calculate the temperature prediction interval based on the real-time inspection data using the temperature prediction model, and judge whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval; as well as, The absolute temperature over-limit and relative temperature over-limit fault judgment module is configured to judge whether the converter valve has an absolute temperature over-limit and relative temperature over-limit fault based on real-time inspection data and using over-temperature judgment criteria.

10. The system according to claim 9, characterized in that: The temperature prediction model construction module obtains the infrared inspection data of the converter valves in each valve hall in the converter station, the operation data of the valve cooling system, the temperature monitoring data of the valve control submodule and the historical data of the operating conditions, and establishes a discrete data set, including, taking the end time of two consecutive infrared scheduled inspections as the start and end intervals of the data set, obtaining the operating status of the infrared inspection system of the converter valves in each valve hall in the converter station within the time period, the operating status of the valve cooling system, the temperature monitoring and operating condition historical information of the valve control system submodule, thereby establishing a discrete data set of characteristic quantities.

11. The system according to claim 9, characterized in that: The temperature prediction model building module establishes a temperature prediction model according to the discrete data set, including: Determine a data set {X1, X2, X3, ...} of input feature parameters and a data set {Y1, Y2, ...} of output feature parameters, perform correlation analysis on the data sets, and use the input feature parameters and output feature parameters that are higher than a set threshold as input data; Preprocessing and standardizing the input data; The processed input data is divided into training set and test set, and convolutional neural network is used for model prediction; The training model uses MSE, RMSE, MAE and R 2 Four evaluation indicators are used to evaluate the model until it is completed as a temperature prediction model; MSE represents mean square error, RMSE represents root mean square error, MAE represents mean absolute error, R 2 represents the coefficient of determination.

12. The system of claim 9, wherein: The submodule temperature fault judgment module judges whether the submodule temperature is normal according to whether the actual temperature of the converter valve submodule falls within the temperature prediction interval, including: After removing abnormal values ​​from the real-time inspection data, the submodule temperature online monitoring system, valve cooling system operating status and operating condition data sets are obtained, and the temperature prediction range of this inspection task is calculated using the temperature prediction model; It is determined whether the actual temperature of the converter valve submodule falls within the temperature prediction interval. If it does not fall within the interval, it is determined that the actual temperature of the converter valve submodule exceeds the predicted temperature rise; otherwise, it is determined that the temperature of the converter valve submodule is normal.

13. The system of claim 9, wherein: The over-temperature criterion building module determines the sub-module component temperature over-temperature alarm set value and the sub-module component relative similar abnormal temperature rise set value, and builds the over-temperature criterion, including the over-temperature criterion including the absolute temperature over-limit criterion and the relative temperature over-limit criterion; The absolute temperature over-limit criterion includes: any point temperature value ≥ M1, which is determined as the absolute temperature over-limit level III alarm of the converter valve; M1>any point temperature ≥ M2, which is determined as the absolute temperature over-limit level II alarm of the converter valve; wherein M1 is the first set value of the submodule component temperature over-temperature alarm, M2 is the second set value of the submodule component temperature over-temperature alarm, and M1>M2; The relative temperature over-limit criterion includes: the temperature difference between submodules ≥ N1, which is judged as the relative temperature over-limit level II alarm of the flow valve submodule; the relative temperature difference of the copper busbar joint of the submodule ≥ N2, which is judged as the relative temperature over-limit level II alarm of the copper busbar joint of the converter valve; wherein N1 is the first fixed value of the abnormal temperature rise of the submodule component relative to the same type, and N2 is the second fixed value of the abnormal temperature rise of the submodule component relative to the same type.

14. The system of claim 9, wherein: The absolute temperature over-limit and relative temperature over-limit fault judgment module uses the over-temperature judgment criteria to judge whether the converter valve has an absolute temperature over-limit, including: Get patrol temperature values ​​in real time; The inspection temperature values ​​are updated in real time through polling and judged one by one.

15. The system of claim 9, wherein: Also includes, The data cleaning module is configured to clean the real-time inspection data before the submodule temperature fault judgment module, the absolute temperature over-limit and the relative temperature over-limit fault judgment module make judgments based on the real-time inspection data, including: Analyze all data points of each camera during one inspection and count the same temperature points. When the number of the same temperature points exceeds half of the total temperature data, it is determined that the PTZ may be faulty. For cameras that are judged to be possibly faulty, the current inspection data is used as the starting point, and three consecutive inspection data are calculated forward to determine whether the number of identical temperature points each time exceeds half of the total temperature data; at the same time, the three inspection data are compared according to the point positions. If they are all the same, it is determined that the infrared pan / tilt head is faulty.

16. The system of claim 9, wherein: Also includes, The fault warning module is configured to generate corresponding alarm fault information according to the judgment results after the submodule temperature fault judgment module makes a judgment and / or after the absolute temperature over-limit and relative temperature over-limit fault judgment modules make a judgment.

17. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor; characterized in that: When the processor executes the computer program, the steps of the method for online temperature monitoring of multiple operating parameters of an MMC converter valve are implemented as described in any one of claims 1 to 8.

18. A computer-readable storage medium storing a computer program; characterized in that: When the computer program is executed by a processor, the steps of the method for online temperature monitoring of multiple operating parameters of an MMC converter valve as described in any one of claims 1 to 8 are implemented.

Images