Radar station lightning protection device online monitoring method and system based on Internet of Things

Abstract

The invention discloses a radar station lightning protection device online monitoring method and system based on the Internet of Things, and relates to the technical field of equipment online monitoring. According to the invention, sensors are deployed in each component of the lightning protection device, real-time transmission and integration of monitoring data are realized by using the Internet of Things technology, a health degree evaluation model based on artificial intelligence is constructed, and remote automatic monitoring and health degree prediction of the lightning protection device of the radar station in a remote area are realized; according to the invention, the health degree evaluation model is combined with the correction coefficient to analyze the assembly evaluation data, and the health degree of each component is accurately predicted, so that the lightning protection effect of the lightning protection device is determined; according to the technical scheme, the space-time limitation of traditional manual inspection is broken through, the problem that a lightning protection device of a radar station in a remote area is difficult to monitor is solved, and conversion from passive maintenance to active prediction is achieved; through real-time data acquisition and intelligent analysis, potential hidden dangers of components can be found in advance, so that workers can take measures in time, and the lightning stroke risk is reduced.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of online monitoring of equipment, and particularly relates to a radar station lightning protection device online monitoring method and system based on Internet of Things. BACKGROUND

[0002] As a key electronic information facility, the radar station is usually deployed in remote areas such as mountain tops and coasts, and its operation stability is extremely vulnerable to lightning threats. The current monitoring and maintenance of the lightning protection device of the radar station faces double technical bottlenecks: on the one hand, the traditional manual inspection mode is limited by the geographical environment of remote areas, and has problems such as long inspection cycle and high labor cost, and the staff need to carry a large number of measuring instruments to collect data on site, and the analysis efficiency of massive monitoring data is low, and it is difficult to find potential hazards of the lightning protection device in real time; on the other hand, the existing automatic monitoring equipment is easily disturbed in extreme environments, and the reliability of sensor data decreases, and it is difficult to continuously and accurately evaluate the health status of each component of the lightning protection device, resulting in a big discount in monitoring effect.

[0003] The present application provides a radar station lightning protection device online monitoring method and system based on Internet of Things, which is used to solve the above technical problems. SUMMARY

[0004] The present application aims to at least solve one of the technical problems existing in the prior art; for this purpose, the present application provides a radar station lightning protection device online monitoring method and system based on Internet of Things, which realizes remote automatic monitoring and health degree prediction of the lightning protection device of the radar station in remote areas by deploying sensors in each component of the lightning protection device, realizing real-time transmission and integration of monitoring data by using Internet of Things technology, and constructing a health degree evaluation model based on artificial intelligence; potential hazards of components can be found in advance, so that staff can take measures in time, reduce the risk of lightning stroke, and ensure the safe operation of the radar station equipment, while greatly reducing the cost of manual inspection and improving the operation and maintenance efficiency.

[0005] To achieve the above purpose, the first aspect of the present application provides a radar station lightning protection device online monitoring method based on Internet of Things, comprising: When the working state of the lightning protection device deployed in the radar station is normal, the monitoring parameters corresponding to the lightning protection device are extracted; the monitoring parameters associated with each component are integrated into component evaluation data; Based on the component evaluation data and the health degree evaluation model, the health degree of each component is predicted; the lightning protection effect of the lightning protection device is determined based on the health degree of each component; wherein the health degree evaluation model is constructed based on an artificial intelligence model; Lightning prediction parameters are extracted; whether the lightning protection device needs to work normally is judged based on the lightning prediction parameters and the lightning protection effect of the lightning protection device; wherein the lightning prediction parameters include lightning intensity and frequency.

[0006] Preferably, the monitoring parameters corresponding to the lightning protection device are extracted, including: Identify the last inspection record, and take the inspection record time as the reference time; Extract the monitoring parameters of the lightning protection device after the reference time; wherein the monitoring parameters refer to the monitoring data affecting the working state of the lightning protection device.

[0007] Preferably, the monitoring parameters associated with each component are integrated into component evaluation data, including: According to the deployment relationship between the monitoring parameters corresponding to the sensors and each component, the monitoring parameters associated with each component are extracted; The monitoring parameters associated with the components are pre-processed and integrated into component evaluation data; wherein the data preprocessing includes time alignment and outlier removal.

[0008] Preferably, the monitoring parameters associated with each component are integrated into component evaluation data, including: According to the deployment relationship between the monitoring parameters corresponding to the sensors and each component, the monitoring parameters associated with each component are extracted as monitoring basic parameters; based on the monitoring basic parameters, the monitoring prediction parameters are obtained by prediction; The basic monitoring parameters and monitoring prediction parameters of the components are spliced and pre-processed, and integrated into component evaluation data; wherein the data preprocessing includes time alignment and outlier removal.

[0009] Preferably, the health degree of each component is predicted, including: The component evaluation data is input into the health degree evaluation model to obtain the score influence degree; Based on the health score of the corresponding component in the last inspection record, a correction coefficient is matched; according to the correction coefficient, the score influence degree is corrected, and the corrected result and the health score are superimposed to obtain the final health score.

[0010] Preferably, based on the health score of the corresponding component in the last inspection record, a correction coefficient is matched, including: Extract the health score of the corresponding component in the last inspection record; Match the correction coefficient corresponding to the health score in the pre-constructed score-correction coefficient table; wherein the score-correction coefficient table is constructed based on simulation experiment.

[0011] Preferably, the health degree of each component is predicted, including: Extract the health score of the corresponding component in the last inspection record; After splicing the health score and the component evaluation data, input them into the health degree evaluation model to obtain the final health score of the component.

[0012] Preferably, whether the lightning protection device needs to work normally is judged based on the lightning prediction parameter and the lightning protection effect of the lightning protection device, including: The lightning prediction parameter is analyzed, and the lightning intensity and the occurrence time are extracted. The lightning protection effect of the lightning protection device is simulated and analyzed to determine whether it can work normally at the occurrence time; if yes, no maintenance treatment is needed; if no, a staff is dispatched to maintain.

[0013] The second aspect of the present application provides an online monitoring system for a lightning protection device of a radar station based on the Internet of Things, which comprises a monitoring and analysis module and a data acquisition module connected thereto. The data acquisition module is used to acquire monitoring parameters of each component of the lightning protection device; wherein the components include a lightning arrester, a down conductor, a grounding module and a surge protector. The monitoring and analysis module is used to integrate the monitoring parameters associated with each component into component evaluation data; based on the component evaluation data and a health degree evaluation model, the health degree of each component is predicted; based on the health degree of each component, the lightning protection effect of the lightning protection device is determined; and, The lightning prediction parameter is extracted; whether the lightning protection device needs to work normally is judged based on the lightning prediction parameter and the lightning protection effect of the lightning protection device; wherein the lightning prediction parameter includes the lightning intensity and frequency.

[0014] Compared with the prior art, the present application has the following advantages: 1. In the present application, sensors are deployed in each component of the lightning protection device, real-time transmission and integration of monitoring data are achieved by using the Internet of Things technology, a health degree evaluation model based on artificial intelligence is constructed, and remote and automatic monitoring and health degree prediction of the lightning protection device of a radar station in a remote area are realized; in the present application, the health degree evaluation model analyzes the component evaluation data in combination with the correction coefficient, accurately predicts the health degree of each component, and then determines the lightning protection effect of the lightning protection device; this technical solution breaks through the time and space limitations of traditional manual inspection, solves the problem of monitoring of the lightning protection device of a radar station in a remote area, and realizes the change from passive maintenance to active prediction; through real-time data acquisition and intelligent analysis, the present application can discover potential hazards of components in advance, so that the staff can take timely measures, reduce the risk of lightning stroke, ensure the safe operation of the radar station equipment, greatly reduce the cost of manual inspection, and improve the operation and maintenance efficiency.

[0015] 2.The application realizes dynamic assessment of the lightning protection device health state and forward-looking maintenance decision by introducing a time series prediction model to predict the future trend of the monitoring parameters, inputting the segmented and spliced current monitoring data and predicted data into the health degree assessment model, and combining lightning prediction parameters; using models such as LSTM to predict the monitoring basic parameters, obtaining monitoring prediction parameters for a future period of time, and splicing the current data to form component assessment data containing time dimension, so that the health degree assessment model can predict the component health degree of each period in the future; at the same time, combined with the lightning intensity and frequency and other prediction parameters obtained from the meteorological platform, the working capacity of the lightning protection device under the future lightning environment can be judged; the technical scheme can adapt to the dynamic influence of extreme environment on the lightning protection device, plan maintenance in advance, avoid sudden failure caused by environmental changes, ensure the normal work of the lightning protection device in the key period, improve the reliability and adaptability of the radar station lightning protection system to extreme environment, realize intelligent dynamic adjustment of operation and maintenance strategy, and effectively ensure the stable operation of the radar station. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1 The method steps of the radar station lightning protection device online monitoring method in the embodiment of the present application are shown in the figure. Figure 2 The generation steps of the component assessment data in the embodiment of the present application are shown in the figure. Figure One Figure 3 The generation steps of the component assessment data in the embodiment of the present application are shown in the figure. Figure Two DETAILED DESCRIPTION

[0018] The technical solutions of the present application will be described in detail below in conjunction with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0019] Embodiment one: Please refer to Figure 1 , the first aspect of the present application provides a radar station lightning protection device online monitoring method based on Internet of Things, comprising: ​​S100: When the lightning protection device deployed at the radar station is in a normal working state, extract the monitoring parameters corresponding to the lightning protection device; integrate the monitoring parameters associated with each component into component evaluation data; S200: Based on the component evaluation data and the health degree evaluation model, predict the health degree of each component; determine the lightning protection effect of the lightning protection device based on the health degree of each component; wherein the health degree evaluation model is based on an artificial intelligence model; S300: Extract lightning prediction parameters; determine whether the lightning protection device needs to work normally based on the lightning prediction parameters and the lightning protection effect of the lightning protection device; wherein the lightning prediction parameters include lightning intensity and frequency.

[0020] As an important electronic information facility, the stability of the radar station is easily affected by lightning. Effective monitoring of the lightning protection device of the radar station can timely detect potential hazards, reduce the risk of lightning strikes, and ensure the safety of the radar station equipment and the stability of data transmission.

[0021] The lightning protection device of the radar station mainly consists of a lightning receptor, a down conductor, a grounding module, and a surge protector. The installation of the lightning protection device is based on the "Code for Design of Building Lightning Protection" (GB50057) and the professional lightning protection standard for radar stations.

[0022] The lightning receptor is installed at the top of high-rise structures such as radar station machine rooms, antenna towers, and masts, ensuring that its height is higher than the protected object. The lightning receptor actively absorbs lightning discharges and directs lightning current into the down conductor, protecting the underlying buildings or equipment from direct lightning strikes.

[0023] The down conductor is connected directly from the lightning receptor to the grounding module along the building's outer wall, with the requirement that the down conductor's laying path be short and straight, avoiding bends and sharp turns, and should be hidden and easy to maintain. The down conductor safely guides the lightning current received by the lightning receptor to the grounding module, reducing the impedance and induction effects of the lightning current during conduction.

[0024] The grounding module is buried around the radar station underground and is laid in a ring or grid shape to expand the grounding area. The grounding module rapidly discharges the lightning current conducted by the down conductor into the ground, reducing the grounding resistance and avoiding ground potential counterattack damage to equipment.

[0025] The surge protector (PSD) is installed at the power supply inlet end of the radar station's main distribution box (primary PSD) to intercept lightning current invading from external power lines; installed in floor distribution boxes or equipment room distribution boxes (secondary PSD) to further limit surge voltage; installed in the power socket or front end of precision equipment (such as radar mainframe, control system, etc.) (tertiary PSD) to protect terminal equipment. The through-flow capacity of primary PSD, secondary PSD, and tertiary PSD decreases in turn.

[0026] The radar station can be arranged in various remote areas, such as mountain tops and coastal areas, due to its wide range of uses, and therefore has a high probability of being exposed to extreme environments, such as strong winds, salt spray, high humidity, and large temperature and humidity changes. Extreme environments can affect the lightning protection device of the radar station, causing the lightning protection effect of the lightning protection device to gradually decrease. If the radar station is arranged in a remote area, it is difficult to monitor the lightning protection device by manual inspection.

[0027] S100: When the lightning protection device deployed at the radar station is normal, the monitoring parameters corresponding to the lightning protection device are extracted; the monitoring parameters associated with each component are integrated into component evaluation data.

[0028] As described above, due to the wide distribution of the construction location of the radar station and the high probability of being arranged in an extreme environment, the same type of lightning protection device for the radar station is affected differently in different extreme environments, and its health status is of course different. Moreover, the functions of each component of the lightning protection device are different, and their ability to adapt to extreme environments is also different, so they may be affected to different degrees in the same extreme environment.

[0029] Considering that if the lightning protection device is monitored by manual inspection, the staff not only needs to carry a large number of measuring instruments to collect monitoring data, but also needs to analyze a large amount of monitoring data to comprehensively evaluate the lightning protection device. Of course, automated monitoring equipment can also be deployed at the radar station, but the automated monitoring equipment will also be affected by the extreme environment, and cannot guarantee the monitoring effect of the lightning protection device.

[0030] It is also worth noting that the current monitoring and evaluation of lightning protection devices mainly compares the collected monitoring data with the set range. If it meets the set range, it can be understood that the data is normal, and if all monitoring data is normal, it can be understood as normal work. However, obviously, normal work at this time does not mean that the lightning protection device has the same lightning protection effect as when it was just out of the factory, and if this is used to determine whether it can resist future lightning, the evaluation result is obviously not accurate enough.

[0031] To solve the above technical problems, sensors are arranged in each component of the lightning protection device, the data collected by the sensors are transmitted to the control center by using the Internet of Things, the control center comprehensively analyzes these data to predict and evaluate the health status of each component of the lightning protection device.

[0032] The lightning receptor is directly struck by lightning, and its overall state is affected by various extreme environments. For example, the instantaneous high temperature of a direct lightning strike can cause the metal on the lightning receptor to melt, and the lightning current sensor can be installed on the body of the lightning receptor; when the wind speed is large, it will generate mechanical stress on the lightning receptor, which may cause the structure to deform, and the vibration acceleration sensor can be installed on the lightning receptor fixing base; under the condition of long-term high temperature and sunlight, the surface temperature of the lightning receptor rises, which can accelerate the oxidation of the metal, and the temperature sensor can be installed at the tip of the lightning receptor.

[0033] The down conductor is used to transmit lightning current, and its overall state is affected by various extreme environments. For example, the temperature of the down conductor will rise suddenly due to the large instantaneous peak value of the lightning current, which may cause the metal to overheat and melt or the insulating layer to burn out, and the lightning current sensor is arranged at the main position of the down conductor; the electromagnetic force generated by the lightning current and strong winds, hail, etc. can cause the deformation or damage of the down conductor, and the strain sensor is installed at the fixed node of the down conductor, and the vibration sensor is installed on the down conductor support or fixing base; humid, acid rain, salt fog, etc. Environment can accelerate the corrosion of the metal of the down conductor, reduce the conductivity and result strength, and the corrosion monitoring sensor is installed on the exposed part of the down conductor (such as a breakage card), to monitor the corrosion degree of the metal surface.

[0034] The grounding module is used to reduce the grounding resistance and ensure efficient discharge of lightning current, and its overall state is affected by various extreme environments. For example, acidic soil (pH <4), saline-alkali land or industrial wastewater infiltration can accelerate the corrosion of the metal module, causing the grounding resistance to rise, and the corrosion monitoring sensor is installed on the surface of the metal module; drought causes the soil to dry and crack, and the grounding module does not contact the soil well; heavy rain and waterlogging may cause soil subsidence and damage the structure of the grounding grid; the instantaneous heating (up to several thousand degrees Celsius) of the lightning current may cause the metal module to melt locally or cause the non-metal module to carbonize and fail; the lightning current impact recorder is installed at the connection between the down conductor and the grounding module; the soil expands when it freezes, which can squeeze the grounding module and cause it to deform, and the soil may collapse after thawing, which can pull apart the connection point; the stress and strain sensor is installed at the fixed node of the grounding module.

[0035] The surge protector is used to limit transient overvoltage and shunt surge current, and its overall state is affected by various extreme environments. For example, frequent lightning current or operating overvoltage impact can cause the MOV grain boundary to deteriorate, the leakage current to increase, and even thermal breakdown, and the internal gas ionization efficiency of the GDT (gas discharge tube) decreases after multiple discharges, and the breakdown voltage drifts, and the leakage current monitoring sensor is installed on the two sections of the MOV element or the grounding loop; high temperature (>70℃) accelerates the chemical decomposition of the MOV inside, and low temperature (<-40℃) causes the breakdown voltage of the GDT to rise; high humidity (>90%RH) may cause internal short circuit or insulation of the SPD to drop, and the temperature sensor is installed near the internal components of the SPD, and the humidity sensor is installed inside the SPD cavity.

[0036] The above lists the environmental factors that may affect the state of each component of the lightning protection device for the radar station, and provides corresponding sensors and installation methods for collecting each environmental factor. The sensors collect monitoring parameters for evaluating the health status of each component of the lightning protection device. After completing data collection, the sensors can send the monitoring parameters to a remote control center through a data transmission device provided in the radar station.

[0037] In S100, the lightning protection device deployed at the radar station is in a normal working state.

[0038] The normal working state of the lightning protection device means that it can work normally and achieve the purpose of lightning protection for the radar station. Whether the lightning protection device is in a normal working state can be determined by maintenance records, which are obtained by personnel manually checking the recorded data, such as checking whether the connections of each component of the lightning protection device are abnormal, and testing the working state of each component and the lightning protection device by instruments.

[0039] It should be noted that the main problem to be solved by the present application is the online remote monitoring of the lightning protection device corresponding to the radar station located in a remote area. The technical solution is to predict the health status of the lightning protection device by transmitting monitoring data through the Internet of Things in most cases. If the health status of the lightning protection device or its components does not meet the requirements, personnel still need to be dispatched to conduct on-site detection of the lightning protection device, which will generate maintenance records or detection records. If the lightning protection device has not been maintained, it is assumed that all components are in a qualified state when leaving the factory, which can be understood as that the health scores of all components are full marks.

[0040] In some other preferred embodiments, it can also be determined whether the lightning protection device or each component thereof can work normally by the data collected by the sensors. For example, if the soil moisture content decreases from 25% to 8% and the grounding body welding point is rusted, it can be determined that the grounding module is abnormal; if the SPD leakage current is 40 μA (close to the threshold value of 50 μA) but does not trigger an alarm, it can be determined that the SPD is abnormal.

[0041] The above only lists the possible environmental influences, and if the components are also affected by other environmental factors in practice, appropriate sensors can be selected for installation and data collection. The data collected by each sensor can be transmitted to a remote control center in real time or at regular intervals through a data transmission module, which can be transmitted centrally or classified.

[0042] In S100, the monitoring parameters corresponding to the lightning protection device are extracted; and the monitoring parameters associated with each component are integrated into component evaluation data.

[0043] Whether the working state of the lightning protection device is determined by manual inspection record or sensor data, only whether the lightning protection device can work normally can be determined, if the lightning protection device can work normally, but the health degree of the lightning protection device and each component thereof is unknown. Especially in extreme environment, although the lightning protection device and each component thereof can work normally, it may be in sub-health state due to the influence of environmental factors, and may fail at any time due to extreme environmental influence. Therefore, it is necessary to determine the influence of extreme environment on each component of the lightning protection device according to the monitoring data, so as to provide a basis for evaluating the lightning protection effect of the lightning protection device.

[0044] In a preferred embodiment, the monitoring parameters corresponding to the lightning protection device are extracted, including: S111: identifying the last inspection record, and taking the inspection record corresponding time as the reference time; S112: extracting the monitoring parameters of the lightning protection device after the reference time; wherein the monitoring parameters refer to the monitoring data affecting the working state of the lightning protection device.

[0045] The last inspection record refers to the record data obtained by dispatching staff to conduct actual inspection according to the warning signal in the remote monitoring process. If the radar station or its lightning protection device is just completed, there is no inspection record, and the completion time is taken as the reference time.

[0046] The monitoring parameters of the lightning protection device are mainly used to evaluate the influence of extreme environment on the health degree of each component, so the monitoring parameters are actually the monitoring data collected by the sensors deployed in the lightning protection device and each component. The sensor type and deployment principle are as described above, and the collected monitoring data mainly include temperature, humidity, deformation, lightning current, corrosion and other data. The monitoring data collected by each component is different.

[0047] In a preferred embodiment, please refer to Figure 2 The monitoring parameters associated with each component are integrated into component evaluation data, including: S121: according to the deployment relationship between the monitoring parameter corresponding sensor and each component, extracting the monitoring parameters associated with each component; S122: data preprocessing of the monitoring parameters associated with each component, and integrating into component evaluation data; wherein the data preprocessing includes time alignment and outlier elimination.

[0048] The deployment relationship between the sensor and the component part is mainly used to determine whether the monitoring parameter and the component part are associated. For example, the sensor is installed on a certain component part and is used to collect the relevant monitoring parameter of the component part, and it is determined that the collected monitoring parameter is associated with the component part. It should be noted that if the sensor is set to collect monitoring data, which can be used to evaluate the health status of each component part but is not set on the component part, it can also be considered that the monitoring data is associated with the component part.

[0049] In this way, starting from the last inspection record (or the completion of construction), the monitoring data of each type corresponding to the component part can be extracted, and the values and time sequence change characteristics of the monitoring data will directly affect the health degree of the component part. The monitoring data of the component part is time-aligned, and the abnormal values are removed and interpolated, and integrated into the component evaluation data, which includes a plurality of types of monitoring data.

[0050] In another preferred embodiment, referring to Figure 3 The monitoring parameters associated with each component part are integrated into the component evaluation data, including: S121: According to the deployment relationship between the sensor corresponding to the monitoring parameter and each component part, the monitoring parameter associated with each component part is extracted as the monitoring basic parameter; and based on the monitoring basic parameter, the monitoring prediction parameter is obtained by prediction; S122: The basic monitoring parameter and the monitoring prediction parameter of the component part are spliced and data preprocessed, and integrated into the component evaluation data; wherein the data preprocessing includes time alignment and abnormal value elimination.

[0051] If the health degree of each component part in the lightning protection device is predicted based on the monitoring parameter of the component part that has been collected, and the lightning protection effect is determined based on the health degree of each component part and is jointly analyzed with the lightning prediction parameter, the influence of the monitoring parameter on the health degree of each component part within the prediction time is ignored. For example, the last inspection record corresponds to the reference time, the monitoring parameter up to the current time is extracted, and the health degree of the component part obtained according to the monitoring parameter is the health degree corresponding to the current time; and the lightning prediction parameter is the lightning parameter at a future time, such as the lightning parameter at the thirtieth day in the future; if the health degree of each component part at the current time is compared with the lightning parameter at the thirtieth day in the future, there is a thirty-day prediction time difference in between, and if an extreme environment appears within the thirty days, the health degree of each component part at the thirtieth day in the future is inconsistent with the current time.

[0052] In order to solve the problem that the health degree and the corresponding time of the lightning prediction parameter do not match, and the comparison between the two may lead to abnormal results, the monitoring parameter of the component part is predicted to the future monitoring parameter, and the current and future monitoring parameters are combined to analyze the health degree of the component part.

[0053] After the monitoring parameters (monitoring basic parameters) of the constituent components are preprocessed, they are input into a pre-trained time series prediction model, such as an LSTM model, to predict the monitoring parameters in the future period of time. The monitoring parameters obtained by the prediction are monitoring prediction parameters. The monitoring prediction parameters and the monitoring basic parameters are spliced, so that the spliced component evaluation data contains future time. When the component evaluation coefficient is input into the health degree evaluation model, the health degree at the current time can be obtained, and the health degree at the future time can also be predicted.

[0054] The LSTM model is disclosed in the prior art for predicting future data based on time series data, and will not be described in detail here.

[0055] It is worth noting that, since the monitoring prediction parameters need to correspond to a large number of monitoring basic parameters of the constituent components, the extraction of the monitoring parameters associated with each constituent component is not limited by the reference time corresponding to the last inspection record.

[0056] S200: predicting the health degree of each constituent component based on the component evaluation data and the health degree evaluation model; and determining the lightning protection effect of the lightning protection device based on the health degree of each constituent component.

[0057] After obtaining the component evaluation coefficient of each constituent component, it is input into the trained health degree evaluation model to obtain the health score corresponding to each constituent component, which is taken as the health degree of the constituent component. After obtaining the health degree of each constituent component of the lightning protection device, the overall lightning protection effect of the lightning protection device can be comprehensively evaluated. The overall lightning protection effect of the lightning protection device refers to the lightning intensity and frequency that can be prevented by the lightning protection device in the current state.

[0058] It should be noted that the component evaluation data of the constituent components is input into the health degree evaluation model, and the health degree at the last collection time of the monitoring parameters in the component evaluation data is output. For example, the monitoring parameters corresponding to the component evaluation data are from ten days ago to yesterday, and the output health degree corresponds to yesterday.

[0059] If the component evaluation data includes monitoring prediction parameters, the monitoring prediction parameters can be segmented and spliced with the monitoring basic parameters. For example, the monitoring basic parameters are up to the current time, and the monitoring prediction parameters are monitoring data in the future one month. The monitoring prediction parameters can be segmented according to every five days to obtain six segments of monitoring prediction parameters. The monitoring basic parameters and the first segment of monitoring prediction parameters are spliced to obtain component evaluation parameter one; the monitoring basic parameters and the first and second segments of monitoring prediction parameters are spliced to obtain component evaluation parameter two; and so on; and the monitoring basic parameters and the first to sixth segments of monitoring prediction parameters are spliced to obtain component evaluation parameter six.

[0060] The segmented and spliced component evaluation data is input into the health assessment model, and the health of the component part every five days in the next month can be obtained. The health curve of the component part in the next month can be obtained by data fitting, and can be called at any time in subsequent analysis of whether the lightning protection device can work normally.

[0061] After determining the lightning protection effect of the lightning protection device according to each component part of the lightning protection device, it can be judged whether the lightning protection effect meets the requirements. If it does not meet the requirements, the staff needs to be dispatched in time to maintain the lightning protection device of the radar station. Similarly, if the component evaluation data is segmented and spliced by monitoring and prediction parameters, it can be judged according to the health curve when the staff needs to be dispatched to the radar station to maintain the lightning protection device.

[0062] In a preferred embodiment, the health of each component is predicted, including: S210: inputting the component evaluation data into the health assessment model to obtain the score influence degree; S220: matching the correction coefficient based on the health score of the corresponding component part in the last inspection record; correcting the score influence degree according to the correction coefficient, and superimposing the correction result and the health score to obtain the final health score.

[0063] The score influence degree refers to the health influence degree of the corresponding component evaluation data on the component part. The health influence degree is relative to the influence degree of the component part in the health state (factory state), so it also needs subsequent matching correction coefficient and matching process according to the correction coefficient.

[0064] The health assessment model of the embodiment is trained based on the test data obtained by simulating the test of the lightning protection component in the factory state. The qualified component parts in the factory are selected, and the corresponding health influence degree of the component part is tested by simulating various extreme environments. A large number of health influence degrees corresponding to the test environment can be obtained; the test monitoring data is used as the input of the artificial intelligence model, and the health influence degree is used as the output of the artificial intelligence model, and the health assessment model is trained. The artificial intelligence model can be an LSTM model or a deep convolutional neural network model. The input data and output data need to be standardized during training, so that the artificial intelligence model can recognize the data.

[0065] After determining the component evaluation data of each component part of the lightning protection device, it is standardized and input into the above-mentioned health assessment model, and the score influence degree corresponding to the component part can be obtained, that is, the health influence degree of the component evaluation data on the component part in the factory qualified state. The score influence degree is not the predicted health score of the component part, and needs to be corrected and superimposed with the health score in the last inspection record to obtain the health score corresponding to the component part.

[0066] In a preferred embodiment, the matching correction coefficient is based on the health score of the corresponding component part in the last inspection record, comprising: S221: extracting the health score of the corresponding component part in the last inspection record; S222: matching the correction coefficient corresponding to the health score in the pre-constructed score-correction coefficient table; wherein the score-correction coefficient table is constructed based on simulation experiments.

[0067] During the last inspection process, the staff will conduct detailed inspection and record of each position of the lightning protection device, and according to the record, the health score of each component part of the lightning protection device can be accurately evaluated. Based on the health score, the correction coefficient corresponding to the health stage can be matched through the score-correction coefficient table, and the correction coefficient is multiplied by the score influence degree and then superimposed with the health score to obtain the health score of the component part under the current state.

[0068] The score-correction coefficient table is also constructed by simulation test. As in the training data of the health degree evaluation model described above, it includes the health influence degree corresponding to several groups of test monitoring data. Test the component parts in different health states using several groups of test monitoring data, and obtain the corresponding health influence degree. By comparing the influence degree with the corresponding health influence degree in the training data, the correction coefficient under the corresponding health state can be obtained. Using multiple health states and correction coefficients, the score-correction coefficient table can be constructed.

[0069] As an example but not limited to, a group of test monitoring data and corresponding health influence degree of a lightning arrester with factory qualified (health score 100 points) is taken as an example. Select lightning arresters with different health scores, including 80, 60, 40, and 20 points. Test the lightning arrester with a health score of 80 points using test monitoring data, and obtain its health influence degree (health score before test minus health score after test). The ratio of the health influence degree to the health influence degree corresponding to the test monitoring data is taken as the correction coefficient. If multiple groups of test monitoring data and corresponding health influence degree are tested, multiple correction coefficients under the same health state can be obtained, and the average of the correction coefficients is taken as the final correction coefficient and is associated with the health score. Based on the health score and the corresponding correction coefficient, the score-correction coefficient table of the lightning arrester can be obtained.

[0070] It should be noted that the health score obtained by multiplying the correction coefficient with the score influence degree and then superimposing with the health score should be less than the health score of the component part corresponding to the last inspection record.

[0071] The health degree evaluation model of the embodiment can only simulate the component part with a health score of 100 during training, and the health degree evaluation model is trained by simulation data. In actual use, the influence degree of the obtained score needs to be corrected in combination with the score-correction coefficient table, and the final health score is obtained in combination with the health score of the component part in the last inspection record. Although it is slightly complex in actual use, the training process is relatively simple, and the required data volume is relatively small.

[0072] S300: Extracting a lightning prediction parameter; determining whether the lightning protection device needs to work normally based on the lightning prediction parameter and the lightning protection effect of the lightning protection device; wherein the lightning prediction parameter includes lightning intensity and frequency.

[0073] The lightning prediction parameter can be obtained by a lightning monitoring and forecasting system of a meteorological platform, such as the lightning intensity and duration of the location of a radar station provided by the China Meteorological Administration; other professional meteorological prediction platforms can also realize lightning prediction.

[0074] In a preferred embodiment, determining whether the lightning protection device needs to work normally based on the lightning prediction parameter and the lightning protection effect of the lightning protection device includes: S311: Analyzing the lightning prediction parameter, extracting the lightning intensity and the occurrence time; S312: Simulating and analyzing whether the lightning protection effect of the lightning protection device can work normally at the occurrence time; if yes, no maintenance treatment is needed; if no, a staff is dispatched for maintenance.

[0075] After the lightning prediction parameter is extracted, it is determined whether the lightning protection effect of the lightning protection device can resist the lightning intensity and frequency in the lightning prediction parameter, if yes, a staff is not needed to maintain the lightning protection device, if no, a staff is needed to immediately maintain the lightning protection device.

[0076] At this time, the lightning protection effect of the lightning protection device at the corresponding time of the lightning prediction parameter needs to be matched. If the time interval between the lightning prediction parameter and the current time is not long, such as 6 hours, the lightning protection effect of the lightning protection device at the current time is used for judgment; if the time interval is long, such as 2 days, the lightning protection effect of the lightning protection device at the corresponding time of the lightning prediction parameter is determined according to the health degree curve of each component part of the lightning protection device.

[0077] In the construction process of the aforementioned component evaluation data, if the monitoring parameters of the constituent parts are integrated into component evaluation data after data preprocessing, and the component evaluation data is input into the health assessment model, the health score of the constituent parts at the current time is obtained, and therefore, when determining whether maintenance is needed, only the lightning in the near future can be analyzed, such as lightning in 6 hours. If the lightning comes before the lightning protection device is subjected to extreme environments, the health score obtained from the component evaluation data is not accurate enough to determine whether the lightning protection device can work normally in a long period of time.

[0078] On the contrary, the component data in the aforementioned scheme can also be integrated and generated according to the basic monitoring parameters and the monitoring prediction parameters, that is, the future environmental changes are predicted according to the currently obtained basic monitoring parameters. After the predicted monitoring prediction parameters are spliced with the basic monitoring parameters, the health score prediction of the constituent parts in the future period of time can be more accurate. After the lightning prediction parameters are determined, the health scores of the constituent parts of the lightning protection device at the corresponding time can be matched according to the time when the lightning occurs, so as to determine whether the lightning protection device can resist the lightning intensity at the corresponding time.

[0079] Embodiment two: Compared with embodiment one, the training process of the health assessment model in this embodiment is different.

[0080] Based on the component evaluation data and the health assessment model, the health of each constituent component is predicted, including: S210: Extracting the health score of the corresponding constituent part in the last inspection record; S220: Splicing the health score with the component evaluation data and inputting it into the health assessment model to obtain the final health score of the constituent part.

[0081] The training of the health assessment model in this embodiment is different from that in embodiment one, mainly in the input data and output data of the model training. Different health scores of the constituent parts are selected, and the health scores of the constituent parts after being subjected to different extreme environments are simulated. The health score of the constituent part before simulation and the test monitoring data are spliced as input data, and the health score after simulation is used as output data. The artificial intelligence model is trained by using the input data and the output data, and the health assessment model is obtained.

[0082] When predicting the health score of each constituent part of the lightning protection device, the health score of the constituent part in the last inspection record can be spliced with the monitoring data after the last inspection to form input data, and the current health score of the constituent part can be obtained by inputting the input data into the health score model.

[0083] The embodiment does not need to be corrected when training the artificial intelligence model, and the health score of the component part in the last inspection record is directly obtained as the model input, thereby obtaining the health score of the component part. Although the training process of the health degree evaluation model is complex, the health degree evaluation model is convenient and high in precision when used. The artificial intelligence model selects the basic structure.

[0084] The second aspect embodiment of the present application provides an online monitoring system for a lightning protection device of a radar station based on the Internet of Things, comprising a monitoring and analysis module, and a data acquisition module connected thereto. The data acquisition module is used to acquire monitoring parameters of each component part of the lightning protection device; wherein the component parts include a lightning arrester, a down conductor, a grounding module and a surge protector. The monitoring and analysis module is used to integrate the monitoring parameters associated with each component part into component evaluation data; predict the health degree of each component based on the component evaluation data and the health degree evaluation model; determine the lightning protection effect of the lightning protection device based on the health degree of each component; and, The data acquisition module is used to acquire monitoring parameters of each component part of the lightning protection device; wherein the component parts include a lightning arrester, a down conductor, a grounding module and a surge protector.

[0085] The data acquisition module and the sensors provided in the lightning protection device perform data transmission through the Internet of Things, and the sensors include lightning current sensors, temperature sensors, humidity sensors, corrosion monitoring sensors, etc. The monitoring and analysis module is mainly responsible for the training and updating of the health degree evaluation model and the scheduling of the health degree evaluation model to evaluate the health degree of the component parts, and determines the maintenance time in combination with the lightning prediction parameters.

[0086] The above embodiments are only used to illustrate the technical method of the present application and are not limited. Although the present application is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. An online monitoring method for radar station lightning protection devices based on the Internet of Things, characterized in that, include: When the lightning protection device deployed at the radar station is working normally, the monitoring parameters corresponding to the lightning protection device are extracted; The monitoring parameters associated with each component are integrated into component evaluation data; Based on the component evaluation data and health assessment model, the health of each component is predicted; the lightning protection effect of the lightning protection device is determined based on the health of each component; wherein, the health assessment model is constructed based on an artificial intelligence model; Extract lightning prediction parameters; determine whether the lightning protection device needs to function properly based on the lightning prediction parameters and the lightning protection effect of the device; among which, the lightning prediction parameters include lightning intensity and frequency.

2. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 1, characterized in that, Extract the monitoring parameters corresponding to the lightning protection device, including: Identify the previous inspection record and use the corresponding time in the inspection record as the baseline time; Extract the monitoring parameters of the lightning protection device after the reference time; wherein, the monitoring parameters refer to the monitoring data that affect the working status of the lightning protection device.

3. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 1, characterized in that, The monitoring parameters associated with each component are integrated into component evaluation data, including: Based on the deployment relationship between the sensors and each component corresponding to the monitoring parameters, the monitoring parameters associated with each component are extracted; The monitoring parameters associated with the components are preprocessed and integrated into component evaluation data; the data preprocessing includes time alignment and outlier removal.

4. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 1, characterized in that, The monitoring parameters associated with each component are integrated into component evaluation data, including: Based on the deployment relationship between the sensors and each component corresponding to the monitoring parameters, the monitoring parameters associated with each component are extracted as the basic monitoring parameters; based on the basic monitoring parameters, prediction is made to obtain the predicted monitoring parameters; The basic monitoring parameters and monitoring prediction parameters of the components are spliced ​​and preprocessed to form component evaluation data; the data preprocessing includes time alignment and outlier removal.

5. A method for online monitoring of radar station lightning protection devices based on the Internet of Things, as described in claim 3 or 4, characterized in that, Predicting the health of each component, including: The component evaluation data is input into the health assessment model to obtain the degree of impact of the score; Based on the health scores of the corresponding components in the previous inspection record, a correction coefficient is matched; the influence of the correction coefficient on the score is corrected, and the correction result is superimposed with the health score to obtain the final health score.

6. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 5, characterized in that, Based on the health score matching correction coefficients of the corresponding components in the previous inspection record, including: Extract the health score of the corresponding component from the previous inspection record; Match the health score to the correction coefficient in a pre-constructed score-correction coefficient table; wherein the score-correction coefficient table is constructed based on a simulation experiment.

7. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 1, characterized in that, Predicting the health of each component, including: Extract the health score of the corresponding component from the previous inspection record; The health score is concatenated with the component evaluation data and then input into the health assessment model to obtain the final health score of the component.

8. The online monitoring method for radar station lightning protection devices based on the Internet of Things according to claim 1, characterized in that, Determining whether a lightning protection device needs to function properly based on lightning prediction parameters and the effectiveness of the lightning protection device includes: Analyze lightning prediction parameters to extract lightning intensity and occurrence time; The simulation analysis determines whether the lightning protection device can function normally at the moment of lightning strike; if yes, no maintenance is required; otherwise, personnel should be dispatched for maintenance.

9. An online monitoring system for radar station lightning protection devices based on the Internet of Things (IoT), used to execute the online monitoring method for radar station lightning protection devices based on the Internet of Things (IoT) as described in any one of claims 1 to 4, characterized in that, It includes a monitoring and analysis module, and a data acquisition module connected to it; Data acquisition module: used to collect monitoring parameters of each component of the lightning protection device; the components include lightning rods, down conductors, grounding modules and surge protectors; Monitoring and Analysis Module: This module integrates monitoring parameters associated with each component into component evaluation data; based on the component evaluation data and a health assessment model, it predicts the health of each component; based on the health of each component, it determines the lightning protection effect of the lightning protection device; and, Used to extract lightning prediction parameters; based on the lightning prediction parameters and the lightning protection effect of the lightning protection device, to determine whether the lightning protection device needs to be able to work normally; among which, the lightning prediction parameters include lightning intensity and frequency.

Images