Bus duct system monitoring method and device, electronic equipment and readable storage medium

Abstract

The invention provides a bus duct system monitoring method and device, electronic equipment and a readable storage medium, and the method comprises a first preparation stage, a second preparation stage and a use stage. The first preparation stage comprises the following steps: acquiring a reference sample of each region; training the initial model by using the reference sample to obtain a target model; the second preparation stage comprises the following steps: dividing the target bus duct system into a plurality of monitored areas; using the position information of each monitored area, the first real environment data and the aging parameters to generate a training sample; training the target model by using the training sample to obtain an available model; the use stage comprises the steps of inputting historical second real environment data at the current moment into the usable model, and obtaining an aging degree prediction result of the external insulation structure in each monitored area; and determining the monitoring mode of each monitored area according to the aging degree prediction result. Through the method, the monitoring accuracy is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of bus ducts, and in particular to a bus duct system monitoring method, device, electronic equipment and readable storage medium. Background Art

[0002] The bus duct system is an important equipment for power transmission and distribution, which is widely used in various industrial and commercial buildings. It is composed of copper or aluminum busbars installed in a closed metal device to achieve efficient and safe power transmission.

[0003] In order to ensure the safety of use, an insulating structure is wrapped around the outside of the bus duct. However, as the use time increases, the insulating structure will be corroded, resulting in a decrease in the overall safety of the bus duct system. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a bus duct system monitoring method, device, electronic device and readable storage medium to ensure the accuracy of monitoring and reduce the time required for training the model when the model is used specifically.

[0005] In a first aspect, an embodiment of the present application provides a bus duct system monitoring method, which acts on a bus duct system arranged in a coastal area; the method comprises: a first preparation stage, a second preparation stage and a use stage;

[0006] The first preparation stage includes:

[0007] Obtain reference samples for each region, each of which is generated using a reference bus duct system located in a different region; the reference samples carry location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times;

[0008] Use the reference samples to train the initial model to obtain the target model;

[0009] The second preparation stage includes:

[0010] Divide the target bus duct system into a plurality of monitored areas according to the environmental information of each location in the target bus duct system;

[0011] For each monitored area, obtaining aging parameters of the external insulation structure in the monitored area at different times and first real environment data of the monitored area at different times;

[0012] Generate training samples using the location information of each monitored area, the first real environment data and the aging parameters;

[0013] Using the training samples to train the target model to obtain a trained and usable model;

[0014] The use phase includes:

[0015] Obtain the second real environment data of each monitored area in real time;

[0016] At predetermined intervals, the historical second real environment data at the current moment is input into the usable model to obtain a prediction result of the aging degree of the external insulation structure in each monitored area;

[0017] The monitoring mode of each monitored area is determined according to the aging degree prediction result.

[0018] In combination with the first aspect, the embodiment of the present application provides a first possible implementation of the first aspect, wherein the step of determining the monitoring mode of each monitored area according to the aging degree prediction result includes:

[0019] Determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession;

[0020] If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

[0021] In combination with the first aspect, an embodiment of the present application provides a second possible implementation of the first aspect, wherein the time span of the data included in the training sample is smaller than the time span of the data in the reference sample.

[0022] In combination with the first aspect, the embodiment of the present application provides a third possible implementation of the first aspect, wherein the initial model is trained using a reference sample to obtain a target model, including:

[0023] Clustering the regions corresponding to the reference samples using the location information, reference aging parameters and reference real environment data in the reference samples to obtain at least one clustering result;

[0024] For each clustering result, the initial model is trained using the reference samples corresponding to the regions in the clustering result to obtain the target model.

[0025] In combination with the third possible implementation of the first aspect, the embodiment of the present application provides a fourth possible implementation of the first aspect, wherein, if the clustering result is greater than or equal to 2, the target model is trained using the training sample to obtain a trained and usable model, including:

[0026] Using the data in the training sample, respectively calculate the similarity between the training sample and each clustering result;

[0027] The target model corresponding to the clustering result with the highest similarity is trained using the training samples to obtain a trained and usable model.

[0028] In a second aspect, an embodiment of the present application further provides a bus duct system monitoring device, which acts on a bus duct system arranged in a coastal area; the operation process of the device includes: a first preparation stage, a second preparation stage and a use stage;

[0029] The device comprises:

[0030] A first preparation module is used to obtain reference samples for each region in a first preparation stage, each of which is generated by using a reference bus duct system located in a different region; the reference sample carries location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times; the initial model is trained using the reference samples to obtain a target model;

[0031] A second preparation module is used to divide the target bus duct system into a plurality of monitored areas according to the environmental information of each position in the target bus duct system in the second preparation stage; for each monitored area, obtain the aging parameters of the external insulation structure in the monitored area at different times and the first real environmental data of the monitored area at different times; generate training samples using the location information of each monitored area, the first real environmental data and the aging parameters; train the target model using the training samples to obtain a trained and usable model;

[0032] The use module is used to obtain the second real environment data of each monitored area in real time during the use phase; at a predetermined interval, the historical second real environment data at the current moment is input into the usable model to obtain the aging degree prediction result of the external insulation structure in each monitored area; and the monitoring method of each monitored area is determined according to the aging degree prediction result.

[0033] In combination with the second aspect, an embodiment of the present application provides a first possible implementation of the second aspect, wherein the use module, when used to determine the monitoring mode of each monitored area according to the aging degree prediction result, is specifically used to:

[0034] Determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession;

[0035] If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

[0036] In combination with the second aspect, an embodiment of the present application provides a second possible implementation of the second aspect, wherein the time span of the data included in the training sample is smaller than the time span of the data in the reference sample.

[0037] In combination with the second aspect, an embodiment of the present application provides a third possible implementation of the second aspect, wherein the first preparation module, when used to train the initial model using the reference sample to obtain the target model, is specifically used to:

[0038] Clustering the regions corresponding to the reference samples using the location information, reference aging parameters and reference real environment data in the reference samples to obtain at least one clustering result;

[0039] For each clustering result, the initial model is trained using the reference samples corresponding to the regions in the clustering result to obtain the target model.

[0040] In combination with the third possible implementation of the second aspect, the embodiment of the present application provides a fourth possible implementation of the second aspect, wherein, if the clustering result is greater than or equal to 2, the second preparation module, when used to train the target model using the training sample to obtain a trained and usable model, is specifically used to:

[0041] Using the data in the training sample, respectively calculate the similarity between the training sample and each clustering result;

[0042] The target model corresponding to the clustering result with the highest similarity is trained using the training samples to obtain a trained and usable model.

[0043] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps in any possible implementation of the first aspect above are performed.

[0044] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in any possible implementation of the first aspect described above are executed.

[0045] The bus duct system monitoring method provided in the embodiment of the present application, in the preparation stage, first uses reference samples located in different regions to train a general model, that is, a target model, and then for the target bus duct system to be monitored, first collects data trip training samples in a short period of time, and then uses the training samples to perform a secondary training on the target model, so that the trained usable model is more suitable for the target bus duct system, and finally uses the trained usable model to collect data from different monitored areas in the target bus duct system, and uses the collected data to monitor, thereby ensuring the accuracy of monitoring and reducing the time required for training the model when the model is specifically used.

[0046] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A flow chart of a bus duct system monitoring method provided by an embodiment of the present application is shown;

[0049] Figure 2 A three-dimensional schematic diagram of a bus duct system provided in an embodiment of the present application is shown;

[0050] Figure 3 A schematic top view of a bus duct system provided in an embodiment of the present application is shown;

[0051] Figure 4 A schematic diagram showing multiple regions provided in an embodiment of the present application;

[0052] Figure 5 A three-dimensional schematic diagram of dividing a bus duct system into regions provided in an embodiment of the present application is shown;

[0053] Figure 6 A schematic top view of a bus duct system divided into regions provided in an embodiment of the present application is shown;

[0054] Figure 7 A flow chart of a second bus duct system monitoring method provided in an embodiment of the present application is shown;

[0055] Figure 8A flow chart of a third bus duct system monitoring method provided in an embodiment of the present application is shown;

[0056] Fig. 9 A schematic diagram of the structure of a bus duct system monitoring device provided in an embodiment of the present application is shown;

[0057] Fig.10 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0058] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0059] In the related art, the outer surface of the bus duct structure in the bus duct system is provided with an insulation structure, which prevents the bus duct from leaking the transmitted current to the outside during operation, causing safety problems or other risks in use. However, as the use time increases, the external insulation structure will age, leading to various problems, such as:

[0060] 1. Increased power loss: As the insulating material ages, its resistivity will increase, resulting in increased power loss during transmission.

[0061] 2. Shortened service life: Aging of insulation materials will accelerate the shortening of the service life of the bus duct, requiring more frequent maintenance and replacement.

[0062] 3. Accelerated equipment aging: Aging insulation materials may cause adjacent electrical equipment, such as wires and cables and electrical insulation supports, to age faster, increasing the risk of system failure.

[0063] 4. Increased fire risk: Aging of insulation materials may reduce their heat resistance and easily cause fire accidents.

[0064] Furthermore, there are many different ways to detect the aging of the external insulation structure in the related art, for example:

[0065] 1. Small current measurement sensor monitoring: Through the small current measurement sensor connected to the busbar grounding wire, the leakage current on the grounding wire can be detected. Once the leakage current is detected, the system will generate and send a signal for further processing and response.

[0066] 2. Data acquisition module analysis: The data acquisition module is responsible for collecting parameters from the small current measurement sensor, such as the leakage current. When the parameters exceed the preset alarm threshold, the module will generate an alarm message and send out these parameters together with the alarm message.

[0067] 3. Wireless data communication module transmission: The wireless data communication module receives the parameters and alarm information from the data acquisition module, and then uploads this information to the monitoring data concentration unit through the wireless network. This wireless transmission method facilitates remote data monitoring in complex or difficult wiring environments.

[0068] 4. Monitoring data centralized unit processing: The monitoring data centralized unit is responsible for receiving the parameters and alarm information uploaded wirelessly, and storing and uploading the alarm information to the background monitoring center. This link ensures the centralized management and rapid response of information.

[0069] 5. Backstage monitoring center management: The backstage monitoring center receives and notifies alarm information, and also provides the function of querying and managing data. This enables operation and maintenance personnel to understand the operating status of the bus duct system in real time and take corresponding measures when necessary.

[0070] However, these monitoring methods are all general monitoring methods, which cannot achieve the effect of advance prediction, resulting in increased monitoring costs and uncontrollable monitoring. In response to the above situation, the present application provides a bus duct system monitoring method, which acts on the bus duct system arranged in the coastal area.

[0071] like Figure 1 As shown, the method comprises: a first preparation stage, a second preparation stage and a use stage;

[0072] The first preparation stage includes:

[0073] S101, obtaining a reference sample for each region, each reference sample is generated using a reference bus duct system located in a different region; the reference sample carries location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times;

[0074] S102, training the initial model using the reference sample to obtain a target model;

[0075] The second preparation phase includes:

[0076] S103, dividing the target bus duct system into a plurality of monitored areas according to the environmental information of each position in the target bus duct system;

[0077] S104, for each monitored area, obtaining aging parameters of the external insulation structure in the monitored area at different times and first real environment data of the monitored area at different times;

[0078] S105, generating training samples using the location information of each monitored area, the first real environment data and the aging parameter;

[0079] S106, training the target model using the training samples to obtain a trained and usable model;

[0080] The use phase includes:

[0081] S107, acquiring the second real environment data of each monitored area in real time;

[0082] S108, at a predetermined time interval, inputting the historical second real environment data at the current moment into the usable model to obtain a prediction result of the aging degree of the external insulation structure in each monitored area;

[0083] S109, determining a monitoring method for each monitored area according to the aging degree prediction result.

[0084] The method provided in the present application is divided into three stages when implemented, namely two preparation stages (the first preparation stage and the second preparation stage) and the use stage. In the first preparation stage, reference samples from different regions need to be acquired in advance. These reference samples reflect the aging of bus duct systems in different coastal areas. For example, different reference samples can come from coastal cities such as Shanghai, Dalian, Qinhuangdao, and Qingdao. Different reference samples can also come from coastal areas of the same city or the same province. In actual operation, considering the large differences in climate and natural conditions at different latitudes, it is preferred to collect samples in different coastal areas or cities in a coastal province.

[0085] The generation process of reference samples is as follows:

[0086] Firstly, for the reference bus duct system in each region, it is necessary to divide the reference bus duct system into regions. Figure 2 A three-dimensional schematic diagram of a bus duct system provided in an embodiment of the present application is shown. Figure 3 A schematic top view of a bus duct system provided in an embodiment of the present application is shown. Figure 2 and Figure 3As shown, generally speaking, the area where a reference bus duct system is located is a factory building or a room of tens or hundreds of square meters. The area where the bus duct system is located is not a fully enclosed area, so the external humid air will contact various parts of the bus duct system. However, since different parts are located in different positions in the area (such as a factory building), the order of air flow is different, so the humid air has different effects on different parts of the bus duct system, so it is necessary to first talk about the reference bus duct system for regional division. As mentioned above, the first basic principle of division is to divide the area according to the order of air circulation. The division mainly considers the order of air circulation (how many seconds after the air entering the factory building passes through the area) and the intensity of air circulation (wind force). The area division can also be further based on environmental information, which can include temperature, humidity, etc. These environmental information will also affect the aging process, so it is necessary to consider these environmental information to determine the final division strategy. Environmental information can also further include the content of substances in the air. In actual operation, the content of substances in the air is not easy to measure. At the same time, the substances carried in the air are most likely the same, so the value of detecting the content of substances is relatively low. Generally speaking, during the whole process of air circulation, if there is a significant change in the content of substances in the air, it is caused by the special production process in the area (it may also be a situation where some substances have leaked). Therefore, generally speaking, in actual operation, the change in the content of substances in the detection control can be used as the basic standard (if the content of substances in the air has changed significantly, it means that there is a substance leak at present, and an alarm should be issued).

[0087] After dividing each bus duct system into zones, the bus ducts in each zone can be monitored separately.

[0088] like Figure 4 As shown, a schematic diagram of multiple regions is shown. For example, each region includes region A, region B, region C, region D, and region E. Each region corresponds to its own bus duct system. Taking region A as an example, Figure 5 A three-dimensional schematic diagram of a bus duct system divided into regions provided in an embodiment of the present application is shown. Figure 6 FIG. 1 shows a schematic top view of a bus duct system divided into regions provided in an embodiment of the present application; Figure 5 and Figure 6 As shown in FIG. 1 , the bus duct system corresponding to region A is divided into multiple regions.

[0089] There are mainly three types of data that need to be monitored in real time, namely, aging degree data, environmental data and time data, or in other words, what needs to be monitored is the aging degree data (i.e., reference aging parameters) of the corresponding external insulation structure and the changes in environmental data at different consecutive moments. Environmental data mainly refers to various parameters in the environment, such as temperature, humidity, temperature changes, humidity changes, etc. These parameters can also cause changes in aging parameters.

[0090] After obtaining the reference samples, the initial model can be trained so that it can learn the aging rules of the bus duct system in each region. Specifically, the training model mainly learns the changes of various aging parameters over time, regional location, and regional location. After learning these situations, the aging trend can be foreseen in advance through the model during the use stage, so as to monitor in advance.

[0091] However, it is not enough to train the model only in the first preparation stage. The main reason is that in the first preparation stage, the model learns the aging law of each region, which reflects the overall situation, so it is not targeted enough. Therefore, this solution adopts the training process of the second preparation stage to solve this problem.

[0092] Specifically, in the second preparation stage, the parameter information of the monitored target bus duct system needs to be obtained, and the target bus duct system is also the object that needs to be monitored in the use stage. The parameter information obtained is similar to the information in the first preparation stage, and the content of the first real environment data in step S104 and the reference real environment data in step S101 are the same, and will not be repeated here.

[0093] Generally speaking, the time span of the data contained in the training sample in step S105 should be much smaller than the time span of the data in the reference sample in step S101. Furthermore, the data density of the data contained in the training sample in step S105 can be smaller than the data density of the data in the reference sample in step S101. This setting method is mainly based on the consideration that the target bus duct system is the bus duct system to be monitored, and the second preparation stage is only equivalent to a simple sampling of the target bus duct system, so that the target model trained in the first preparation stage is more compatible with the target bus duct system, and the aging of the external insulation structure of the target bus duct system can be better predicted. Therefore, when the target model obtained in step S102 can already reflect the overall aging of the external insulation structure of the bus duct as a whole, there is no need for too many training samples for refined training.

[0094] Among them, the time span of the data contained in the training sample in step S105 should be much shorter than the time span of the data in the reference sample in step S101, which mainly means that the time for collecting the first real environment data is shorter. If it takes several months to collect the reference real environment data, then the time for collecting the first real environment data may be only one week. The shorter data collection time can enable the usable model in step S106 to be trained faster, thereby ensuring that the model can be used earlier.

[0095] The data density of the data contained in the training sample in step S105 may be less than the data density of the data in the reference sample in step S101, which means that the collection frequency of the first real environment data may be lower than the collection time of the reference real environment data. This can reduce the cost of data collection. In other words, the second preparation stage and the use stage may not necessarily be a sequential relationship. In some cases, after the second preparation stage is executed for a period of time (the first period), the first real environment data and aging parameters collected in the first period are used to train the target model. After the first period, the first real environment data and aging parameters can continue to be collected for a period of time (the second period). After the second period is collected, the first real environment data and aging parameters collected in the second period can be used to further optimize the usable model. Specifically, the collection of the second period and the acquisition after the collection can be repeated to repeatedly optimize the usable model, thereby ensuring that the usable model can be adapted in real time according to the actual situation of the target bus duct system.

[0096] Afterwards, in the use phase, the second real environment data acquired in real time can be input into the usable model, and then the aging degree prediction results for each monitoring area in the target bus duct system can be obtained. Then, in step S109, corresponding monitoring strategies can be formulated for different monitored areas according to the prediction results of the aging degree. Generally speaking, the higher the aging degree of the monitored area, the more accurate the monitoring measures should be, the higher the monitoring frequency should be, and correspondingly, the higher the monitoring cost will be.

[0097] In specific implementation, since in the first preparation stage, the reference samples are the detection results of reference bus duct systems in multiple different regions, the trained model may be unstable, that is, the prediction results output by the model have certain fluctuations. For this situation, the solution provided in the present application may not directly execute step S109 after obtaining the aging degree prediction result, but in step S109, first detect the stability of the prediction result, and then determine whether to use the prediction result for subsequent monitoring based on the stability.

[0098] Specifically, Figure 7As shown, step S109 can be implemented as follows:

[0099] S1091, determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession;

[0100] S1092: If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

[0101] In step S1091, the aging degree prediction results obtained multiple times in succession are generally the prediction results continuously output by the model. The first real environment data used in these multiple times of aging degree prediction results are different. Specifically, this requires step S108 to be repeated multiple times, rather than just once. Each time step S108 is executed, the data used is also different. Each time step S108 is executed, the first real environment data used are the data obtained up to the current moment.

[0102] The first way of using data: for example, executing step S108 10 times in total, thereby obtaining ten groups of prediction results (each group of prediction results includes the aging degree prediction results corresponding to different monitored areas predicted at the same time), and the first real environment data used by the first to tenth groups of prediction results are the first real environment data detected in 0-5 hours, 0-7 hours, 0-10 hours, 0-13 hours, 0-17 hours, 0-22 hours, 0-28 hours, 0-35 hours, 0-43 hours, and 0-50 hours, respectively.

[0103] The second way of using data: In addition to the above-mentioned way of using the first real environment data, the first real environment data can also be used in the following way, that is, the first real environment data used in the first to tenth groups of prediction results are respectively the first real environment data detected in 0-5 hours, 6-10 hours, 11-15 hours, 16-20 hours, 21-25 hours, 26-30 hours, 31-35 hours, 36-40 hours, 41-45 hours, and 46-50 hours.

[0104] The difference between the above two methods of utilizing the first real environment data is whether the first real environment data used later contains the first real environment data that has been acquired historically. In actual use, if the first real environment data used later contains too much historical data (the first real environment data used by the previous group of aging degree prediction results), it may cause the part different from the historical data (the part of the first real environment data used by the current group of aging degree prediction results that is different from the first real environment data used by the previous group of aging degree prediction results, that is, the newly added part) to be submerged in the historical data, thereby making the judgment of the degree of stability made in step S1091 not accurate enough. Therefore, when using the above first method of utilizing data, it should be ensured that at least 20%-30% of the first real environment data of the current group should be different from the historical data.

[0105] Generally speaking, the sampling time of the first real environment data used in a set of prediction results is suitable to be controlled between nearly one hour and several hours. If the sampling time is too short, the instability of the environment may easily cause data instability, which may lead to the failure of the judgment of step S1091. If the sampling time is too long, the external insulation structure may not be able to update the monitoring strategy in time after aging, which may lead to the occurrence of hidden dangers.

[0106] In step S1092, if the degree of stability meets the preset requirements (there is no obvious difference between multiple aging degree prediction results), the aging degree prediction results can be used to determine the monitoring method for each monitored area. The specific process of determining the monitoring method is the same as that described in the previous content and will not be repeated here.

[0107] As described above, step S102 is the process of training the model for the first time. The training performed in this step can make the model a model with strong generalization ability, and it is also the basis for the following steps. Since the training of the model in the second preparation stage only plays the role of accuracy adjustment, it is more important whether the model can be effectively and fully trained in the first stage.

[0108] Specifically, there are two ways to perform step S102. The first way is to train a comprehensive model in step S102, then collect samples in the second preparation stage, and train again to adjust the accuracy of the target model. The second way is to cluster samples from different regions in step S102, and then train the results of each cluster separately, so that the target model trained in the first preparation stage has strong pertinence.

[0109] like Figure 8 As shown, under the second method of executing step S102, step S102 can be implemented as follows:

[0110] S1021, clustering the regions corresponding to the reference samples by using the location information, reference aging parameters and reference real environment data in the reference samples to obtain at least one clustering result;

[0111] S1022: For each clustering result, the initial model is trained using reference samples corresponding to the regions in the clustering result to obtain a target model.

[0112] In step S1021, the main purpose of clustering the reference samples is to merge regions of the same type. As mentioned above, the different regions in this solution are all coastal regions, which can refer to multiple different coastal cities or different regions in coastal cities. The environmental data in each region is affected by various factors (such as differences in terrain, altitude, etc.), and there may be certain differences. Therefore, it is necessary to use clustering to fuse samples of the same type in different regions, and then use the fused results separately for training (only one model is trained for samples in each type of region), so that the trained model is more targeted.

[0113] In step S1021, when clustering is performed, the reference real environment data is mainly used. The more similar the reference real environment data of two regions are, the more likely they are to be classified into the same category. The role of the region is mainly to play an auxiliary role. Generally speaking, the more similar the regions are, the more likely they are to be classified into the same category. The region here may only contain basic location information, and may further contain other information related to regional characteristics, such as altitude, flatness of terrain, climate parameters and weather parameters, etc., all of which will affect the corrosion of the external insulation structure (at least the speed of corrosion), but the actual impact of these parameters on the corrosion situation has not been fully quantified, so they can only be used as auxiliary judgment parameters. In other words, in the process of clustering, the weight of the reference real environment data is greater than, or even much greater than, the weight of the location information (generally speaking, the weight of the reference real environment data can be several times the weight of the location information).

[0114] More specifically, generally speaking, when clustering using only the reference real environment data cannot produce more than or equal to 2 categories, the location information will be used for auxiliary verification. If the clustering situation of the location information is similar to the clustering situation of the reference real environment data, then it can be confirmed that the result of using only the reference real environment data is credible. Otherwise, other auxiliary data should be added for clustering.

[0115] Generally, when clustering, the number of cluster centers should not be specified. Considering the situation of abnormal data, K-means algorithm can be used for clustering first.

[0116] After clustering is completed, in step S1022, each clustering result can be used to train the corresponding initial model, thereby obtaining multiple target models. Each clustering result corresponds to a target model, so that the target model itself can be guaranteed to have a certain degree of specificity.

[0117] The corresponding step S106 can be performed as follows:

[0118] S1061, using the data in the training sample, respectively calculating the similarity between the training sample and each clustering result;

[0119] S1062: Use the training samples to train the target model corresponding to the clustering result with the highest similarity, so as to obtain a trained and usable model.

[0120] In step S1061, the data in the training sample obtained in step S105 can be used to calculate the similarity, with the purpose of selecting the target model that best suits the target bus duct system from all the target models. Specifically, the target model that best suits the target bus duct system is the target model corresponding to the data in the clustering result that is most similar to the data in the training sample. Therefore, the target model with the highest similarity can be selected from all the target models to execute step S1062 by calculating the similarity (or calculating the distance between the training sample and each clustering result).

[0121] In the specific implementation, step S1062 is implemented by using training samples to train the target model corresponding to the clustering result with the highest similarity. This is mainly because even if clustering is performed, it can be guaranteed to a certain extent that the target model corresponding to the clustering result is the most friendly target model to the target bus duct system among all target models, but in essence, the two are still not completely corresponding, so further training can be used in step S1062. However, it should be known that the training process in step S1062 can be completed faster, and further, the process can be accelerated by adjusting the convergence conditions of the training.

[0122] In addition to using steps S1061-S1062 to perform step S106, the following method can also be used to implement step S106:

[0123] Step 1063, using the data in the training sample, respectively calculating the similarity between the training sample and each clustering result;

[0124] Step 1064: Use the target model corresponding to the clustering result with the highest similarity between the training sample and the training sample as the usable model.

[0125] That is, after similarity calculation, the target model corresponding to the clustering result with the highest similarity can be directly used as the usable model without training. This processing method can save the time of secondary training.

[0126] In specific implementation, in the second preparation stage, after collecting the first real environment data, the first real environment data and the reference real environment data contained in the clustering result with the highest similarity are compared. If the similarity between the two is high, steps 1063-1064 can be executed. If the similarity is not high enough, steps S1061-S1062 are executed.

[0127] Based on the same technical concept, the embodiment of the present application also provides a bus duct system monitoring device, which acts on a bus duct system arranged in a coastal area; the operation process of the device includes: a first preparation stage, a second preparation stage and a use stage; Fig. 9 As shown, the device comprises:

[0128] The first preparation module 901 is used to obtain reference samples for each region in the first preparation stage, each of which is generated by using a reference bus duct system located in a different region; the reference sample carries location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times; the initial model is trained using the reference samples to obtain a target model;

[0129] The second preparation module 902 is used to divide the target bus duct system into multiple monitored areas according to the environmental information of each position in the target bus duct system in the second preparation stage; for each monitored area, obtain the aging parameters of the external insulation structure in the monitored area at different times and the first real environmental data of the monitored area at different times; generate training samples using the location information of each monitored area, the first real environmental data and the aging parameters; train the target model using the training samples to obtain a trained and usable model;

[0130] Use module 903, for obtaining the second real environment data of each monitored area in real time during the use phase; input the historical second real environment data of the current moment into the usable model at a predetermined interval to obtain the aging degree prediction result of the external insulation structure in each monitored area; determine the monitoring method of each monitored area according to the aging degree prediction result.

[0131] Optionally, when the using module 903 is used to determine the monitoring mode of each monitored area according to the aging degree prediction result, it is specifically used to:

[0132] Determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession;

[0133] If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

[0134] Optionally, the time span of the data included in the training sample is smaller than the time span of the data in the reference sample.

[0135] Optionally, when the first preparation module 901 is used to train the initial model using the reference sample to obtain the target model, it is specifically used to:

[0136] Clustering the regions corresponding to the reference samples using the location information, reference aging parameters and reference real environment data in the reference samples to obtain at least one clustering result;

[0137] For each clustering result, the initial model is trained using the reference samples corresponding to the regions in the clustering result to obtain the target model.

[0138] Optionally, if the clustering result is greater than or equal to 2, the second preparation module 902, when used to train the target model using the training sample to obtain a trained and usable model, is specifically used to:

[0139] Using the data in the training sample, respectively calculate the similarity between the training sample and each clustering result;

[0140] The target model corresponding to the clustering result with the highest similarity is trained using the training samples to obtain a trained and usable model.

[0141] Fig.10 A structural diagram of an electronic device provided for an embodiment of the present application includes: a processor 1001, a memory 1002 and a bus 1003, wherein the memory 1002 stores machine-readable instructions executable by the processor 1001. When the electronic device runs the above-mentioned information processing method, the processor 1001 communicates with the memory 1002 through the bus 1003, and the processor 1001 executes the machine-readable instructions to perform the method steps described in Example 1.

[0142] The embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method steps described in the first embodiment are executed.

[0143] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, electronic devices, and computer-readable storage media can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0144] In the several embodiments provided in the present application, it should be understood that the disclosed methods, devices, electronic devices and computer-readable storage media can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or modules can be electrical, mechanical or other forms.

[0145] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0146] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0147] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0148] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A bus duct system monitoring method, characterized in that: Acting on a bus duct system arranged in a coastal area; the method comprises: a first preparation stage, a second preparation stage and a use stage; The first preparation stage includes: Obtain reference samples for each region, each of which is generated using a reference bus duct system located in a different region; the reference samples carry location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times; Use the reference samples to train the initial model to obtain the target model; The second preparation stage includes: Divide the target bus duct system into a plurality of monitored areas according to the environmental information of each location in the target bus duct system; For each monitored area, obtaining aging parameters of the external insulation structure in the monitored area at different times and first real environment data of the monitored area at different times; Generate training samples using the location information of each monitored area, the first real environment data and the aging parameters; Using the training samples to train the target model to obtain a trained and usable model; The use phase includes: Obtain the second real environment data of each monitored area in real time; At predetermined intervals, the historical second real environment data at the current moment is input into the usable model to obtain a prediction result of the aging degree of the external insulation structure in each monitored area; The monitoring mode of each monitored area is determined according to the aging degree prediction result.

2. The method according to claim 1, characterized in that: The step of determining the monitoring method of each monitored area according to the aging degree prediction result includes: Determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession; If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

3. The method according to claim 1, characterized in that: The time span of the data contained in the training sample is smaller than the time span of the data in the reference sample.

4. The method according to claim 1, characterized in that: The initial model is trained using reference samples to obtain the target model, including: Clustering the regions corresponding to the reference samples using the location information, reference aging parameters and reference real environment data in the reference samples to obtain at least one clustering result; For each clustering result, the initial model is trained using the reference samples corresponding to the regions in the clustering result to obtain the target model.

5. The method according to claim 4, characterized in that: If the clustering result is greater than or equal to 2, the target model is trained using the training sample to obtain a trained and usable model, including: Using the data in the training sample, respectively calculate the similarity between the training sample and each clustering result; The target model corresponding to the clustering result with the highest similarity is trained using the training samples to obtain a trained and usable model.

6. A bus duct system monitoring device, characterized in that: Acts on bus duct systems installed in coastal areas; The operation process of the device includes: a first preparation stage, a second preparation stage and a use stage; The device comprises: A first preparation module is used to obtain reference samples for each region in a first preparation stage, each of which is generated by using a reference bus duct system located in a different region; the reference sample carries location information of the corresponding region, reference aging parameters of the external insulation structure of each tested area in the reference bus duct system of the corresponding region at different times, and reference real environment data of each tested area in the corresponding region at different times; the initial model is trained using the reference samples to obtain a target model; A second preparation module is used to divide the target bus duct system into a plurality of monitored areas according to the environmental information of each position in the target bus duct system in the second preparation stage; for each monitored area, obtain the aging parameters of the external insulation structure in the monitored area at different times and the first real environmental data of the monitored area at different times; generate training samples using the location information of each monitored area, the first real environmental data and the aging parameters; train the target model using the training samples to obtain a trained and usable model; The use module is used to obtain the second real environment data of each monitored area in real time during the use phase; at a predetermined interval, the historical second real environment data at the current moment is input into the usable model to obtain the aging degree prediction result of the external insulation structure in each monitored area; and the monitoring method of each monitored area is determined according to the aging degree prediction result.

7. The device according to claim 6, characterized in that: When the using module is used to determine the monitoring mode of each monitored area according to the aging degree prediction result, it is specifically used to: Determining the stability of the current prediction based on the aging degree prediction results obtained multiple times in succession; If the stability level meets the preset requirements, the monitoring method of each monitored area is determined according to the aging level prediction result.

8. The device according to claim 6, characterized in that: The time span of the data contained in the training sample is smaller than the time span of the data in the reference sample.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the method as claimed in any one of claims 1 to 5 are performed.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are executed.

Images