Monitoring and protection methods and platforms for the safe operation of power distribution cabinets

Abstract

The present application provides a monitoring and protection method and platform for the safe operation of distribution cabinets, which relates to the field of distribution cabinet protection technology, including: receiving the service cross-sectional data of the first component of the distribution cabinet; performing mechanical life analysis based on the environmental humidity cross-sectional data and the mechanical motion cross-sectional data; performing electrical life analysis based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data; when the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold, or / and the electrical life margin prediction value, generating scrap identification information for the first component of the distribution cabinet, and generating a scrap signal for the associated distribution cabinet; sending the scrap identification information and the scrap signal to the distribution cabinet operation and maintenance management end. The present application can solve the technical problem in the prior art that the hidden faults of the equipment cannot be predicted in advance due to the lack of real-time monitoring, realize the technical goal of real-time monitoring and dynamic prediction of the status of the distribution cabinet, and achieve the technical effect of timely discovering potential hidden dangers.

Description

Technical Field

[0001] The present application relates to the field of power distribution cabinet protection technology, and in particular to a monitoring and protection method and platform for the operation safety of power distribution cabinets. Background Art

[0002] During the operation and maintenance of the distribution cabinet, the long-term use of the equipment will cause the mechanical and electrical components to gradually age, thus affecting the reliability and safety of the equipment. As an important equipment in the power system, the normal operation of the distribution cabinet is directly related to the stability and safety of the entire power network.

[0003] However, in the prior art, there are still many defects in the operation status monitoring and life management of the distribution cabinet. The traditional operation safety protection method of the distribution cabinet mainly relies on the analysis of the maintenance data after the fault occurs, that is, after the equipment has a problem, the operation and maintenance personnel judge the status of the equipment through the fault record and maintenance history data. Although this method can provide assistance for troubleshooting to a certain extent, it is difficult to achieve real-time monitoring and preventive maintenance due to its excessive reliance on post-analysis, and it is impossible to discover potential hidden dangers in time, which often leads to the equipment failure. It is already in an irreversible state. Among them, the components in the distribution cabinet, such as switches, contactors and wires, will gradually age with long-term operation and changes in environmental conditions, and traditional technologies can only repair and replace these components when they have obvious faults, and cannot predict the remaining life of these components in advance. Once a component suddenly fails during operation, it may cause a more serious power accident or even paralyze the entire power grid. Especially when the environmental humidity, temperature and load fluctuate greatly, these external factors will accelerate the aging of the equipment, but traditional technologies do not have effective means to dynamically monitor the changes of these factors, so they cannot accurately predict the true state of the equipment. In addition, traditional fault records can only reflect past problems and cannot capture hidden fault signals generated during equipment operation, such as minor mechanical wear and poor electrical contact. These problems may not appear as obvious faults in the short term, but may lead to serious accidents after long-term accumulation.

[0004] In summary, the existing technology has technical problems such as over-reliance on post-fault maintenance data for analysis and lack of real-time monitoring, which makes it impossible to effectively predict hidden equipment failures and component aging in advance, further affecting the equipment's life management and operation and maintenance efficiency, and increasing the risk of distribution cabinet accidents. Summary of the invention

[0005] The purpose of this application is to provide a monitoring and protection method and platform for the safe operation of distribution cabinets, so as to solve the technical problems in the prior art that due to excessive reliance on post-fault maintenance data for analysis and lack of real-time monitoring, hidden failures of equipment and aging of components cannot be effectively predicted in advance, which further affects the life management and operation and maintenance efficiency of the equipment and increases the risk of accidents in distribution cabinets.

[0006] In view of the above problems, the present application provides a monitoring and protection method and platform for the safe operation of distribution cabinets.

[0007] In the first aspect, the present application provides a monitoring and protection method for the operation safety of a distribution cabinet, which is implemented through a monitoring and protection platform for the operation safety of a distribution cabinet, including: receiving service cross-sectional data of a first component of the distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical movement cross-sectional data and electrical parameter cross-sectional data; performing a mechanical life analysis based on the environmental humidity cross-sectional data and the mechanical movement cross-sectional data to obtain a mechanical life remainder prediction value; performing an electrical life analysis based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life remainder prediction value; when the mechanical life remainder prediction value is less than or equal to the mechanical life remainder threshold, or / and the electrical life remainder prediction value, generating scrap identification information for the first component of the distribution cabinet, and generating a scrap signal for the associated distribution cabinet; and sending the scrap identification information and the scrap signal to the operation and maintenance management end of the distribution cabinet.

[0008] In the second aspect, the present application also provides a monitoring and protection platform for the safe operation of distribution cabinets, which is used to execute the monitoring and protection method for the safe operation of distribution cabinets as described in the first aspect, including: a service section data receiving module, used to receive the service section data of the first component of the distribution cabinet, wherein the service section data includes environmental humidity section data, mechanical movement section data and electrical parameter section data; a mechanical life analysis module, used to perform mechanical life analysis based on the environmental humidity section data and the mechanical movement section data, and obtain a mechanical life margin prediction value; an electrical life analysis module, used to perform electrical life analysis based on the environmental humidity section data and the electrical parameter section data, and obtain an electrical life margin prediction value; a life margin prediction module, used to generate scrap identification information for the first component of the distribution cabinet, and generate a scrap signal for the associated distribution cabinet when the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold, or / and the electrical life margin prediction value; a scrap identification information sending module, used to send the scrap identification information and the scrap signal to the distribution cabinet operation and maintenance management end.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] By receiving the service cross-sectional data of the first component of the distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical movement cross-sectional data and electrical parameter cross-sectional data; performing mechanical life analysis according to the environmental humidity cross-sectional data and the mechanical movement cross-sectional data to obtain a mechanical life margin prediction value; performing electrical life analysis according to the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value; when the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold, or / and the electrical life margin prediction value, generating scrap identification information for the first component of the distribution cabinet, and generating a scrap signal for the associated distribution cabinet; sending the scrap identification information and the scrap signal to the operation and maintenance management end of the distribution cabinet, achieving the technical goal of real-time monitoring of the operating status of key components of the distribution cabinet and dynamic life prediction, and achieving the technical effect of timely discovering potential hidden dangers, predicting component aging conditions in advance and effectively extending the service life of equipment.

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0013] Figure 1 This is a flow chart of the monitoring and protection method for the safe operation of the power distribution cabinet in this application;

[0014] Figure 2 This is a schematic diagram of the structure of the monitoring and protection platform for the safe operation of the distribution cabinet in this application.

[0015] Description of reference numerals:

[0016] A service section data receiving module 11, a mechanical life analysis module 12, an electrical life analysis module 13, a life margin prediction module 14, and a scrap identification information sending module 15. DETAILED DESCRIPTION

[0017] This application provides a monitoring and protection method and platform for the safe operation of distribution cabinets, which solves the technical problem that the existing technology cannot effectively predict the hidden failures of equipment and component aging in advance due to excessive reliance on maintenance data after failures and lack of real-time monitoring, further affecting the life management and operation and maintenance efficiency of equipment, and increasing the risk of distribution cabinet accidents. The technical goal of achieving real-time monitoring of the operating status of key components of distribution cabinets and dynamic life prediction is achieved, achieving the technical effect of timely discovering potential hidden dangers, predicting component aging in advance, and effectively extending the service life of equipment.

[0018] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0019] For example, please refer to the attached Figure 1 The present application provides a monitoring and protection method for the operation safety of a power distribution cabinet, which is applied to a monitoring and protection platform for the operation safety of a power distribution cabinet, and specifically includes:

[0020] Receive service cross-sectional data of a first component of a power distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical motion cross-sectional data, and electrical parameter cross-sectional data.

[0021] Specifically, the service cross-sectional data of the first component of the distribution cabinet is received. The service cross-sectional data refers to various parameter information closely related to the operating status of the first component. For example, the environmental humidity cross-sectional data represents the humidity of the environment where the distribution cabinet is located. Excessive humidity causes electrical equipment to become damp, thereby increasing the risk of failure. Secondly, the mechanical movement cross-sectional data refers to the movement of the mechanical parts in the distribution cabinet, such as the opening and closing actions of the switches in the distribution cabinet. Frequent operation of the switches may increase mechanical wear. Finally, the electrical parameter cross-sectional data refers to the current, voltage and other parameters of the electrical parts in the distribution cabinet. The current and voltage can help to promptly detect possible overload or short circuit risks. Through the comprehensive analysis of the service cross-sectional data, the operating status of the distribution cabinet can be effectively monitored and protective measures can be provided to ensure its safe operation.

[0022] A mechanical life analysis is performed based on the environmental humidity cross-sectional data and the mechanical motion cross-sectional data to obtain a predicted value of the mechanical life margin.

[0023] Specifically, the remaining life of the equipment is predicted by analyzing the humidity of the environment in which the equipment is located and the mechanical movement state of the equipment. By combining these two sets of data, the service life analysis model evaluates the impact of the equipment's working conditions on its life. Ultimately, the predicted value of the mechanical life margin is obtained, and the time the equipment can still operate safely is predicted, which helps to predict the maintenance and replacement time of the equipment, thereby improving the safety and reliability of the equipment.

[0024] An electrical life analysis is performed based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value.

[0025] Specifically, the remaining life of the electrical system of the equipment is predicted by analyzing the humidity and electrical performance indicators of the environment in which the equipment is located. The environmental humidity cross-sectional data refers to the humidity level of the environment when the equipment is running. For example, when the environmental humidity is 80%, higher humidity may cause the electrical components to get damp, which in turn affects the insulation performance and electrical stability of the equipment. The electrical parameter cross-sectional data refers to the electrical operating status of the equipment, including indicators such as current and voltage. For example, when the current is 150A and the voltage is 380V, long-term high current may cause electrical components to overheat or overload, thereby accelerating aging. By combining these two sets of data, the electrical life analysis model is used to evaluate the impact of operating conditions on the life of the electrical system. Finally, the electrical life margin prediction value is obtained to predict the time when the electrical system can still operate safely. For example, the analysis results may show that under the conditions of 80% humidity and 150A current, the remaining life of the electrical system of the equipment is 500 hours, which can effectively predict the maintenance and replacement time of the electrical system and ensure the electrical safety and stability of the equipment.

[0026] When the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold value, or / and the electrical life margin prediction value, scrap identification information is generated for the first component of the distribution cabinet, and a scrap signal is generated for the associated distribution cabinet.

[0027] Specifically, when the remaining mechanical life time obtained through life analysis is lower than the set safety threshold, for example, the remaining life prediction value is 50 hours, and the threshold is set to 100 hours, it indicates that the mechanical part of the equipment is close to the limit of use. At the same time, the life analysis of the electrical system is performed. If the remaining electrical life time is also lower than the threshold, for example, the remaining electrical life is 80 hours, and the threshold is 100 hours, it means that the electrical part also needs to be replaced. When any one or both items are lower than their thresholds, scrap identification information is generated for the first component of the distribution cabinet, and scrap information is automatically generated for the specific component of the distribution cabinet, indicating that the component needs to be replaced or scrapped. The distribution cabinet will generate a scrap signal because the key components are approaching or reaching the end of their life, thereby notifying the maintenance personnel to take corresponding measures. For example, if the mechanical life and electrical life are both lower than their respective thresholds at the same time, the scrap signal is triggered, reminding to replace the distribution cabinet components in time or stop using the distribution cabinet, effectively avoiding the risk of failure caused by excessive use of the equipment and ensuring the safety of equipment operation.

[0028] The scrapping identification information and the scrapping signal are sent to the distribution cabinet operation and maintenance management terminal.

[0029] Specifically, the scrap identification information and the scrap signal are sent to the power distribution cabinet operation and maintenance management end. When a component of the power distribution cabinet reaches its service life limit and generates scrap identification information, the scrap identification information and the scrap signal will be transmitted to the management end responsible for the maintenance of the power distribution cabinet and delivered to the equipment maintenance personnel to help the maintenance personnel arrange equipment repair or replacement in time to avoid sudden failures of the power distribution cabinet. For example, when the management end receives a scrap signal of a power distribution cabinet, it can arrange for shutdown and maintenance to ensure the normal operation and safe operation of the power distribution cabinet.

[0030] The monitoring and protection method for the safe operation of the distribution cabinet is applied to the monitoring and protection platform for the safe operation of the distribution cabinet, which can achieve the technical goal of real-time monitoring of the operating status of key components of the distribution cabinet and dynamic prediction of life, and achieve the technical effect of timely discovering potential hidden dangers, predicting the aging status of components in advance and effectively extending the service life of the equipment.

[0031] Furthermore, the present application also includes:

[0032] The environmental humidity cross-sectional data includes environmental humidity time series recording data, and the mechanical motion cross-sectional data includes motion cycle recording data; the environmental humidity time series recording data is aggregated in the time domain to obtain environmental humidity aggregated time series data; through the mechanical life analysis channel associated with the first component model of the distribution cabinet, the environmental humidity aggregated time series data and the motion cycle recording data are subjected to mechanical life analysis to obtain the predicted value of the mechanical life margin.

[0033] Specifically, the environmental humidity cross-section data includes environmental humidity time series recording data, and the humidity conditions in the working environment of the power distribution cabinet are continuously recorded in chronological order. The environmental humidity time series recording data can be monitored in real time by sensors. For example, the humidity is recorded every hour during a day. The recording results may show that the humidity is 75% in the morning, rises to 85% at noon, and drops to 70% in the evening. The change trend of humidity can be clearly observed through time series recording, so as to determine whether the environmental conditions affect the safe operation of the power distribution cabinet.

[0034] Next, the mechanical motion section data includes motion cycle record data, which tracks and records the operating cycle of mechanical components in the distribution cabinet. For example, the motion cycle record data of the switch in the distribution cabinet can record in detail the time interval and frequency of each opening and closing of the switch. For example, if 100 opening and closing actions occur within two hours, it indicates that there is an abnormality in the distribution cabinet load or other systems that need to be repaired.

[0035] The monitored environmental humidity data is processed to obtain the aggregated time series data of environmental humidity. Time domain hierarchical aggregation refers to the hierarchical arrangement and merging of the data in the time series, integrating the humidity data of multiple time periods into an aggregate value of a larger time period. For example, if the humidity is recorded once every hour, the humidity data within 6 hours can be aggregated into an average value through aggregation operation, thereby reducing the amount of data and highlighting the overall trend. For example, if the humidity within 6 hours is 75%, 76%, 77%, 75%, 74% and 73% respectively, the aggregation result is 75%.

[0036] Through the mechanical life analysis channel associated with the first component model of the power distribution cabinet, the environmental humidity aggregated time series data and the motion cycle record data are input together for mechanical life analysis to obtain the mechanical life margin prediction value. The mechanical life analysis channel can predict the remaining life of the equipment based on the working environment (such as humidity) and working status (such as mechanical motion cycle) of the equipment. For example, if the humidity aggregated data is 75% and the motion cycle is 50 times per hour, the remaining life of the equipment is 300 hours through the analysis channel.

[0037] The environmental humidity data is simplified and integrated through time-domain hierarchical aggregation, and then the humidity data is combined with the mechanical motion cycle data and input into the life analysis model to ultimately derive the remaining mechanical life of the equipment. This allows efficient use of data from different sources, accurately predicts the service life of the equipment, and helps maintenance personnel rationally arrange equipment maintenance and replacement cycles.

[0038] Furthermore, the present application also includes:

[0039] Step 1: Extract the first time domain ambient humidity and the second time domain ambient humidity which are adjacent time domains of the ambient humidity time series recording data; Step 2: When the ambient humidity deviation between the first time domain ambient humidity and the second time domain ambient humidity is less than or equal to the ambient humidity deviation threshold, merge the first time domain and the second time domain into a third time domain; Step 3: Perform mean analysis on the first time domain ambient humidity and the second time domain ambient humidity to obtain the third time domain ambient humidity; Continue to execute steps 1 to 3 based on the third time domain ambient humidity, and when the ambient humidity deviation between any two time domains is greater than the ambient humidity deviation threshold, output the ambient humidity aggregated time series data.

[0040] Specifically, in step 1, the first time domain ambient humidity and the second time domain ambient humidity of the ambient humidity time series recording data, which are adjacent time domains, are extracted, that is, the humidity data of two time periods adjacent in time are selected from the recorded humidity data for comparison. For example, if the humidity recorded at 10:00 and 11:00 is 75% and 76% respectively, then these two humidity data are regarded as the adjacent first time domain ambient humidity and second time domain ambient humidity, and then the humidity changes in adjacent time periods can be gradually compared to find trends or abnormalities.

[0041] In step 2, when the environmental humidity deviation between the first time domain environmental humidity and the second time domain environmental humidity is less than or equal to the environmental humidity deviation threshold, the first time domain and the second time domain are merged into the third time domain, that is, if the humidity difference between the two time periods is within the preset threshold range, it is considered that the environmental conditions of the two time periods are basically the same, so they can be merged into a new time domain. For example, if the humidity difference is only 1%, which is much less than the set 2% threshold, the two time periods from 10 to 11 can be merged into a new time period, namely the third time domain, thereby simplifying the data and reducing unnecessary details, so as to more effectively analyze humidity changes.

[0042] In step 3, the humidity data of the first two time periods are averaged to obtain a new humidity value. For example, the humidity at 10 o'clock is 75%, and the humidity at 11 o'clock is 76%. The proportion of each time period in the merged time period is calculated as the weight of each time period, and the weighted result is calculated as the ambient humidity of the third time domain after the merger, which can smooth out the fluctuations in a short period of time and make the data more representative.

[0043] Based on the ambient humidity in the third time domain, continue to execute steps one to three. Steps one to three are a cyclic process, that is, after each merging, continue to look for the humidity data of the next adjacent time period, and repeat the above comparison and merging operations. The merging operation is not stopped until the humidity deviation between two time periods is greater than the set threshold. For example, if the humidity difference between the third time domain and the fourth time domain is 3%, which exceeds the 2% threshold, then no merging is performed at this time, but the humidity data of these time periods are output as ambient humidity aggregated time series data, and then through a series of merging and mean analysis, simplified and continuous humidity time series data are finally obtained to help obtain the overall trend of humidity changes.

[0044] The complexity of data analysis can be simplified by gradually comparing and merging humidity data from adjacent time periods. By setting a humidity deviation threshold, the data can be effectively prevented from being too dispersed while maintaining data accuracy.

[0045] Furthermore, the present application also includes:

[0046] Calculate the ratio of the first time domain to the third time domain and set it as the first weight; calculate the ratio of the second time domain to the third time domain and set it as the second weight; calculate the weighted mean of the first time domain ambient humidity and the second time domain ambient humidity according to the first weight and the second weight to obtain the third time domain ambient humidity.

[0047] Specifically, the length of the first time domain is compared with the length of the third time domain after merging, and the ratio between the two is calculated. Assuming that the length of the first time domain is 1 hour, and the length of the third time domain is the total length after merging, which is 2 hours, then the ratio is 1:2, so the first weight is 0.5, which prepares for the subsequent weighted average calculation to ensure that the influence of each time domain is in proportion to its length.

[0048] Next, the ratio of the second time domain to the third time domain is calculated and set as the second weight. Assuming that the duration of the second time domain is also 1 hour, and the total duration of the third time domain is still 2 hours, the ratio is also 1:2, so the second weight is also 0.5, which can ensure that the influence of humidity data of different time periods on the result is based on their relative duration when merging.

[0049] Then, the humidity values ​​of the two time periods are weighted averaged according to their respective weights. For example, if the humidity in the first time domain is 75% and the humidity in the second time domain is 76%, and the weighted calculation is performed according to the weight 0.5, the humidity in the third time domain is the sum of 75% multiplied by 0.5 and 76% multiplied by 0.5, and the result is 75.5%. By using the weighted average method, the actual humidity of the combined time period can be more accurately reflected, especially when the time period lengths are inconsistent, it can more fairly represent the humidity conditions of the entire time period.

[0050] By calculating the ratio of the durations and assigning corresponding weights, and then performing weighted mean calculation, it can be ensured that the final humidity value is not just a simple average value, but reflects the relative importance of each time period in the total time, making the calculation result more representative and accurate.

[0051] Furthermore, the present application also includes:

[0052] A mechanical life analysis subnetwork is obtained; the environmental humidity aggregation time series data includes environmental humidity aggregation data of the first time zone until environmental humidity aggregation data of the Nth time zone; according to the environmental humidity aggregation data of the first time zone until the environmental humidity aggregation data of the Nth time zone, the motion cycle recording data is split to obtain the motion cycle recording data of the first time zone until the motion cycle recording data of the Nth time zone; N mechanical life analysis subnetworks are connected in series to generate the mechanical life analysis channel; the first time zone environmental humidity aggregation data, the first time zone motion cycle recording data and the initial mechanical life are input into the first mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the first time zone mechanical life margin prediction value; until the N-1th time zone environmental humidity aggregation data, the N-1th time zone motion cycle recording data and the N-1th time zone mechanical life margin prediction value are input into the Nth mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the mechanical life margin prediction value.

[0053] Specifically, a mechanical life analysis subnetwork is obtained to analyze the life of mechanical equipment. Life assessment is performed based on input data such as environmental humidity and mechanical motion information.

[0054] Next, the environmental humidity aggregated time series data includes the environmental humidity aggregated data of the first time zone to the environmental humidity aggregated data of the Nth time zone, that is, the time period is divided into multiple time zones, each time zone has corresponding environmental humidity data, and the environmental humidity data is obtained through the humidity aggregation step. For example, from 8 am to 4 pm is divided into multiple time zones, each time zone has an aggregated humidity data, such as the humidity of the first time zone is 75%, the humidity of the second time zone is 78%, and so on, and then the humidity information is managed in segments for subsequent analysis.

[0055] According to the environmental humidity aggregate data of the first time zone to the environmental humidity aggregate data of the Nth time zone, the motion cycle record data is correspondingly split to obtain the motion cycle record data of the first time zone to the motion cycle record data of the Nth time zone. For example, if the motion cycle of the mechanical equipment between 8 o'clock and 9 o'clock is 50 times, and the motion cycle between 9 o'clock and 10 o'clock is 60 times, the motion cycle data of each time zone will be matched with the corresponding humidity data, thereby ensuring the synchronous processing of the environment and motion information of each time zone.

[0056] Connect N mechanical life analysis sub-networks in series, connect multiple mechanical life analysis sub-networks in sequence to form a complete analysis channel, and generate a mechanical life analysis channel for analyzing the life of mechanical equipment in different time periods. For example, the overall life analysis process of a device can be composed of multiple steps, each step corresponds to a time zone, and gradually predicts the remaining life of the device.

[0057] The first time zone environmental humidity aggregate data, the first time zone motion cycle record data and the initial mechanical life are input into the first mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the first time zone mechanical life margin prediction value. For example, if the initial life of the equipment is 500 hours, after the analysis of the first time zone, the predicted remaining life may be reduced to 480 hours.

[0058] Until the N-1 time zone environmental humidity aggregation data, the N-1 time zone motion cycle record data and the N-1 time zone mechanical life margin prediction value are input into the N-1 mechanical life analysis subnetwork of the mechanical life analysis channel, the mechanical life margin prediction value is obtained, that is, the data of each time zone is input into the corresponding subnetwork in turn until the last time zone. For example, after analyzing multiple time zones, the overall remaining life of the mechanical equipment is finally obtained, assuming that the final prediction value is 350 hours.

[0059] By establishing multiple life analysis sub-networks and based on the ambient humidity and mechanical movement data of each time zone, the life of mechanical equipment is gradually predicted. The analysis of each time zone will be carried out on the basis of the previous time zone, thus forming a serial analysis process, which can predict the service life of the equipment more accurately and dynamically, and ensure its safe operation in different environments.

[0060] Furthermore, the present application also includes:

[0061] Configure a BP neural network topology, wherein the BP neural network topology includes three input nodes and one output node, the three input nodes are used to input the ambient humidity, motion cycle and initial mechanical life of the monitoring time zone, and the one output node is used to output the predicted value of the mechanical life margin; configure an MIV variable evaluation space, construct a number of hyperparameter particles of the BP neural network topology, wherein any one of the hyperparameter particles includes the hyperparameter assignment results of all nodes of the BP neural network topology; retrieve the historical monitoring data of the first component model of the distribution cabinet, traverse the several hyperparameter particles, verify the BP neural network topology M times, and obtain several particle output mean errors; when the output mean error of any one of the several particle output mean errors is less than or equal to the output mean error threshold, generate the mechanical life analysis subnetwork.

[0062] Specifically, the BP neural network topology is used for prediction or classification tasks. Configure the BP neural network topology, which includes three input nodes. The three input nodes represent different data sources, including the monitored time zone ambient humidity, the movement cycle of the mechanical equipment, and the initial mechanical life of the mechanical equipment. The data of the three input nodes are used together to predict the output, that is, the predicted value of the mechanical life margin of the equipment. For example, assuming that the humidity is 75%, the movement cycle is 50 times per hour, and the initial life is 500 hours, the neural network predicts the remaining mechanical life under the current situation based on the input.

[0063] Configure the MIV variable evaluation space and construct several hyperparameter granules of the BP neural network topology, where any hyperparameter granule includes the hyperparameter assignment results of all nodes of the BP neural network topology, that is, several hyperparameter granules determine the structure and performance of the neural network, such as learning rate, number of hidden layer nodes, and activation function. The MIV variable evaluation space refers to the use of the MIV (mean influence value) method to evaluate the contribution of input variables. Each hyperparameter granule represents a complete hyperparameter configuration scheme for the BP neural network topology, which may contain multiple different settings. For example, a hyperparameter granule may define a learning rate of 0.01, a number of hidden layer nodes of 10, an activation function of ReLU, etc.

[0064] Retrieve the historical monitoring data of the first component model of the power distribution cabinet, traverse several hyperparameter particles, verify the BP neural network topology M times, and obtain the mean error of several particle outputs, that is, use actual historical data to verify the prediction performance of the BP neural network. By verifying different hyperparameter particles multiple times, using the same data set each time, train and test the network to obtain the error between the predicted result and the actual value. For example, if there is a data set containing 1,000 hours of equipment operation data, perform 100 training and verifications on different hyperparameter particles to ensure that the optimal configuration is found.

[0065] When the output mean error of any particle among the output mean errors of several particles is less than or equal to the output mean error threshold, the mechanical life analysis subnetwork is generated, that is, when the prediction error of the hyperparameter particle is found to be less than the preset threshold, the network configuration is determined and the mechanical life analysis subnetwork is generated. For example, if the output mean error threshold is set to 5% and the error of the hyperparameter particle is 4.8%, the network with this configuration is considered to be accurate enough and the corresponding analysis subnetwork is generated.

[0066] By using the MIV variable evaluation space to determine the network's hyperparameter configuration, traversing different hyperparameter combinations, and performing multiple verifications with actual historical data, the mean error is calculated. When the configured error meets the requirements, a mechanical life analysis subnetwork is finally generated to accurately predict the life of mechanical equipment and ensure that the performance under specific conditions reaches the expected goals.

[0067] Furthermore, the present application also includes:

[0068] When the output mean error of any one of the plurality of particle output mean errors is greater than the output mean error threshold, a particle distance evaluation function is obtained:

[0069] ;in, Characterizes the distance between the first hyperparameter particle and the second hyperparameter particle, The first hyperparameter granule, is the second hyperparameter particle, Characterize the result of the i-th dimension assignment of the first hyperparameter particle, Characterizes the second hyperparameter particle i-th dimension assignment result, a represents a small constant, Characterize the i-th dimension hyperparameter weight; group the several hyperparameter particles according to the particle distance evaluation function and in combination with the particle distance threshold to obtain multiple groups of hyperparameter particles; extract multiple head hyperparameter particles of the multiple groups of hyperparameter particles, guide the multiple groups of hyperparameter particles to perform particle distance reduction mutation, and obtain expanded hyperparameter particles; retrieve the historical monitoring data of the first component model of the distribution cabinet, traverse the expanded hyperparameter particles, and verify the BP neural network topology.

[0070] Specifically, when the output mean error of any particle among several particle output mean errors is greater than the output mean error threshold, the particle distance evaluation function is obtained. That is, when the prediction errors of all hyperparameter particles are greater than the preset error threshold, it is necessary to evaluate the similarity of different hyperparameter particles by calculating the distance between particles. The particle distance evaluation function is used to quantify the difference between two different hyperparameter configurations and further optimize the model configuration. For example, if the error threshold is set to 5%, but the errors of all hyperparameter particles exceed 5%, it is necessary to use the distance evaluation function to further refine the analysis.

[0071] in, Characterizes the distance between the first hyperparameter particle and the second hyperparameter particle, is the first hyperparameter particle, is the second hyperparameter particle, Characterize the result of the i-th dimension assignment of the first hyperparameter particle, Characterizes the second hyperparameter particle i-th dimension assignment result, a represents a small constant, Characterizes the weight of the hyperparameter in the i-th dimension. Each hyperparameter particle consists of multiple dimensions (such as learning rate, number of nodes, etc.). The distance is calculated based on the difference in the values ​​assigned to each dimension and weighted. a is a constant used to avoid the denominator being equal to 0, thereby preventing zero values ​​from appearing during calculation. Different weights are assigned to different dimensions of the hyperparameter. For example, the learning rate may be more important than the number of nodes, so the weight will be larger.

[0072] The distance value calculated by the particle distance evaluation function is combined with the preset particle distance threshold to group the hyperparameter particles. The closer the distance between particles in each group is, the similar the hyperparameter configurations are. Assuming the particle distance threshold is set to 0.1, two particles with a distance less than 0.1 will be grouped together, which effectively reduces the computational complexity and allows similar configurations to be processed together.

[0073] The top (i.e., better performing) hyperparameter particles are selected from each group, and further mutation operations are performed to narrow the differences between particles and generate new hyperparameter combinations, that is, by adjusting the hyperparameter values ​​of each dimension to explore a better configuration. For example, the original learning rate of 0.01 can be adjusted to 0.015 during the mutation process to obtain better performance.

[0074] Retrieve the historical monitoring data of the first component model of the power distribution cabinet, traverse the expanded hyperparameter particles, and verify the BP neural network topology, that is, use the expanded hyperparameter particles to verify the BP neural network again, and conduct multiple tests through historical data to evaluate whether the mutated hyperparameter configuration can achieve better prediction results. For example, after expansion and verification, the error of the hyperparameter particles is reduced to 4.5%, which meets the error threshold requirement.

[0075] The similarity between hyperparameter particles is calculated through the particle distance evaluation function, and similar particles are grouped. After that, the head particles are selected from each group and mutated to generate new hyperparameter combinations, expanding the original hyperparameter space to find the optimal hyperparameter configuration and ensure that the neural network can more accurately predict the life of mechanical equipment.

[0076] In summary, the monitoring and protection method for the safe operation of the power distribution cabinet provided by this application has the following technical effects:

[0077] By receiving the service cross-sectional data of the first component of the distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical movement cross-sectional data and electrical parameter cross-sectional data; performing mechanical life analysis according to the environmental humidity cross-sectional data and the mechanical movement cross-sectional data to obtain a mechanical life margin prediction value; performing electrical life analysis according to the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value; when the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold, or / and the electrical life margin prediction value, generating scrap identification information for the first component of the distribution cabinet, and generating a scrap signal for the associated distribution cabinet; sending the scrap identification information and the scrap signal to the operation and maintenance management end of the distribution cabinet, achieving the technical goal of real-time monitoring of the operating status of key components of the distribution cabinet and dynamic life prediction, and achieving the technical effect of timely discovering potential hidden dangers, predicting component aging conditions in advance and effectively extending the service life of equipment.

[0078] Embodiment 2: Based on the monitoring and protection method for the safe operation of the power distribution cabinet in the previous embodiment, the present application also provides a monitoring and protection platform for the safe operation of the power distribution cabinet, please refer to the attached Figure 2 ,include:

[0079] A service cross-sectional data receiving module 11 is used to receive the service cross-sectional data of the first component of the distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical movement cross-sectional data and electrical parameter cross-sectional data; a mechanical life analysis module 12 is used to perform mechanical life analysis based on the environmental humidity cross-sectional data and the mechanical movement cross-sectional data to obtain a mechanical life margin prediction value; an electrical life analysis module 13 is used to perform electrical life analysis based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value; a life margin prediction module 14 is used to generate scrap identification information for the first component of the distribution cabinet and generate a scrap signal for the associated distribution cabinet when the mechanical life margin prediction value is less than or equal to the mechanical life margin threshold, or / and the electrical life margin prediction value; a scrap identification information sending module 15 is used to send the scrap identification information and the scrap signal to the distribution cabinet operation and maintenance management terminal.

[0080] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0081] The environmental humidity cross-sectional data includes environmental humidity time series recording data, and the mechanical motion cross-sectional data includes motion cycle recording data; the environmental humidity time series recording data is aggregated in the time domain to obtain environmental humidity aggregated time series data; through the mechanical life analysis channel associated with the first component model of the distribution cabinet, the environmental humidity aggregated time series data and the motion cycle recording data are subjected to mechanical life analysis to obtain the predicted value of the mechanical life margin.

[0082] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0083] Step 1: Extract the first time domain ambient humidity and the second time domain ambient humidity which are adjacent time domains of the ambient humidity time series recording data; Step 2: When the ambient humidity deviation between the first time domain ambient humidity and the second time domain ambient humidity is less than or equal to the ambient humidity deviation threshold, merge the first time domain and the second time domain into a third time domain; Step 3: Perform mean analysis on the first time domain ambient humidity and the second time domain ambient humidity to obtain the third time domain ambient humidity; Continue to execute steps 1 to 3 based on the third time domain ambient humidity, and when the ambient humidity deviation between any two time domains is greater than the ambient humidity deviation threshold, output the ambient humidity aggregated time series data.

[0084] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0085] Calculate the ratio of the first time domain to the third time domain and set it as the first weight; calculate the ratio of the second time domain to the third time domain and set it as the second weight; calculate the weighted mean of the first time domain ambient humidity and the second time domain ambient humidity according to the first weight and the second weight to obtain the third time domain ambient humidity.

[0086] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0087] A mechanical life analysis subnetwork is obtained; the environmental humidity aggregation time series data includes environmental humidity aggregation data of the first time zone until environmental humidity aggregation data of the Nth time zone; according to the environmental humidity aggregation data of the first time zone until the environmental humidity aggregation data of the Nth time zone, the motion cycle recording data is split to obtain the motion cycle recording data of the first time zone until the motion cycle recording data of the Nth time zone; N mechanical life analysis subnetworks are connected in series to generate the mechanical life analysis channel; the first time zone environmental humidity aggregation data, the first time zone motion cycle recording data and the initial mechanical life are input into the first mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the first time zone mechanical life margin prediction value; until the N-1th time zone environmental humidity aggregation data, the N-1th time zone motion cycle recording data and the N-1th time zone mechanical life margin prediction value are input into the Nth mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the mechanical life margin prediction value.

[0088] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0089] Configure a BP neural network topology, wherein the BP neural network topology includes three input nodes and one output node, the three input nodes are used to input the ambient humidity, motion cycle and initial mechanical life of the monitoring time zone, and the one output node is used to output the predicted value of the mechanical life margin; configure an MIV variable evaluation space, construct a number of hyperparameter particles of the BP neural network topology, wherein any one of the hyperparameter particles includes the hyperparameter assignment results of all nodes of the BP neural network topology; retrieve the historical monitoring data of the first component model of the distribution cabinet, traverse the several hyperparameter particles, verify the BP neural network topology M times, and obtain several particle output mean errors; when the output mean error of any one of the several particle output mean errors is less than or equal to the output mean error threshold, generate the mechanical life analysis subnetwork.

[0090] Furthermore, the mechanical life analysis module 12 in the monitoring and protection platform for the safe operation of the power distribution cabinet is also used for:

[0091] When the output mean error of any one of the plurality of particle output mean errors is greater than the output mean error threshold, a particle distance evaluation function is obtained:

[0092] ;in, Characterizes the distance between the first hyperparameter particle and the second hyperparameter particle, The first hyperparameter granule, is the second hyperparameter particle, Characterize the result of the i-th dimension assignment of the first hyperparameter particle, Characterizes the second hyperparameter particle i-th dimension assignment result, a represents a small constant, Characterize the i-th dimension hyperparameter weight; group the several hyperparameter particles according to the particle distance evaluation function and in combination with the particle distance threshold to obtain multiple groups of hyperparameter particles; extract multiple head hyperparameter particles of the multiple groups of hyperparameter particles, guide the multiple groups of hyperparameter particles to perform particle distance reduction mutation, and obtain expanded hyperparameter particles; retrieve the historical monitoring data of the first component model of the distribution cabinet, traverse the expanded hyperparameter particles, and verify the BP neural network topology.

[0093] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The monitoring and protection method for the safe operation of the distribution cabinet and the specific examples in the aforementioned embodiment 1 are also applicable to the monitoring and protection platform for the safe operation of the distribution cabinet in this embodiment. Through the aforementioned detailed description of the monitoring and protection method for the safe operation of the distribution cabinet, those skilled in the art can clearly understand the monitoring and protection platform for the safe operation of the distribution cabinet in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the platform disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0094] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0095] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A monitoring and protection method for the safe operation of a power distribution cabinet, characterized in that: include: Receive service cross-sectional data of a first component of a power distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical motion cross-sectional data, and electrical parameter cross-sectional data, wherein the environmental humidity cross-sectional data represents the humidity of the environment in which the power distribution cabinet is located, the mechanical motion cross-sectional data refers to the motion of the mechanical part in the power distribution cabinet, and the electrical parameter cross-sectional data refers to the current and voltage parameters of the electrical part in the power distribution cabinet; Performing mechanical life analysis based on the environmental humidity cross-sectional data and the mechanical motion cross-sectional data to obtain a mechanical life margin prediction value; Perform electrical life analysis based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value; When the mechanical life margin prediction value is less than the mechanical life margin threshold, and when the electrical life margin prediction value is less than the electrical life margin threshold, generating scrap identification information for the first component of the power distribution cabinet, and generating a scrap signal for the associated power distribution cabinet; Sending the scrapping identification information and the scrapping signal to the power distribution cabinet operation and maintenance management terminal; The mechanical life analysis is performed according to the environmental humidity cross-sectional data and the mechanical motion cross-sectional data to obtain a predicted value of the mechanical life margin, including: The environmental humidity cross-sectional data includes environmental humidity time series recording data, and the mechanical motion cross-sectional data includes motion cycle recording data; Performing time-domain hierarchical aggregation on the environmental humidity time series record data to obtain environmental humidity aggregated time series data; Through the mechanical life analysis channel associated with the first component model of the power distribution cabinet, a mechanical life analysis is performed on the ambient humidity aggregation time series data and the motion cycle record data to obtain the mechanical life margin prediction value; Performing time-domain hierarchical aggregation on the ambient humidity time series record data to obtain ambient humidity aggregated time series data includes: Step 1: extracting the first time domain ambient humidity and the second time domain ambient humidity of the ambient humidity time series recording data which are adjacent to each other; Step 2: When the ambient humidity deviation between the first time domain ambient humidity and the second time domain ambient humidity is less than or equal to an ambient humidity deviation threshold, merging the first time domain and the second time domain into a third time domain; Step 3: performing mean analysis on the first time-domain ambient humidity and the second time-domain ambient humidity to obtain a third time-domain ambient humidity; Based on the third time-domain ambient humidity, continue to execute steps 1 to 3, and if the deviation of any two time-domain ambient humidity is greater than the ambient humidity deviation threshold, output the ambient humidity aggregated time series data; Performing mean analysis on the first time-domain ambient humidity and the second time-domain ambient humidity to obtain a third time-domain ambient humidity includes: Calculating a ratio of the first time domain to the third time domain and setting the ratio as a first weight; Calculating a ratio of the second time domain to the third time domain and setting the ratio as a second weight; Calculating the weighted mean of the first time-domain ambient humidity and the second time-domain ambient humidity according to the first weight and the second weight to obtain the third time-domain ambient humidity; Among them, through the mechanical life analysis channel associated with the first component model of the power distribution cabinet, the mechanical life analysis is performed on the ambient humidity aggregation time series data and the motion cycle record data to obtain the mechanical life margin prediction value, including: Obtain mechanical life analysis subnetwork; The ambient humidity aggregated time series data includes the ambient humidity aggregated data of the first time zone to the ambient humidity aggregated data of the Nth time zone; According to the first time zone environmental humidity aggregate data up to the Nth time zone environmental humidity aggregate data, split the exercise cycle record data to obtain the first time zone exercise cycle record data up to the Nth time zone exercise cycle record data; Connecting N mechanical life analysis sub-networks in series to generate the mechanical life analysis channel; Inputting the first time zone environmental humidity aggregated data, the first time zone motion cycle recorded data and the initial mechanical life into the first mechanical life analysis subnetwork of the mechanical life analysis channel to obtain a first time zone mechanical life margin prediction value; Until the Nth time zone environmental humidity aggregate data, the Nth time zone motion cycle record data and the N-1th time zone mechanical life margin prediction value are input into the Nth mechanical life analysis subnetwork of the mechanical life analysis channel to obtain the mechanical life margin prediction value; The construction steps of the mechanical life analysis subnetwork are as follows: Configure a BP neural network topology, wherein the BP neural network topology includes three input nodes and one output node, the three input nodes are used to input the ambient humidity of the monitoring time zone, the motion cycle and the initial mechanical life, and the one output node is used to output the predicted value of the mechanical life margin; Configure the MIV variable evaluation space to construct a number of hyperparameter particles of the BP neural network topology, wherein any hyperparameter particle includes the hyperparameter assignment results of all nodes of the BP neural network topology; Retrieving historical monitoring data of the first component model of the power distribution cabinet, traversing the plurality of hyperparameter particles, verifying the BP neural network topology M times, and obtaining a plurality of particle output mean errors; When the output mean error of any one of the plurality of particle output mean errors is less than or equal to the output mean error threshold, the mechanical life analysis sub-network is generated.

2. The method according to claim 1, characterized in that Also includes: When the output mean error of any one of the plurality of particle output mean errors is greater than the output mean error threshold, a particle distance evaluation function is obtained: ， in, Characterizes the distance between the first hyperparameter particle and the second hyperparameter particle, is the first hyperparameter particle, is the second hyperparameter particle, Characterize the result of the i-th dimension assignment of the first hyperparameter particle, Characterizes the second hyperparameter particle i-th dimension assignment result, a represents a small constant, Characterize the hyperparameter weight of the i-th dimension; According to the particle distance evaluation function and in combination with a particle distance threshold, the plurality of hyperparameter particles are grouped to obtain a plurality of groups of hyperparameter particles; Extracting a plurality of head hyperparameter particles from the plurality of groups of hyperparameter particles, guiding the plurality of groups of hyperparameter particles to perform particle distance reduction mutation, and obtaining expanded hyperparameter particles; The historical monitoring data of the first component model of the power distribution cabinet is retrieved, the expanded hyperparameter particles are traversed, and the BP neural network topology is verified.

3. The monitoring and protection platform for the safe operation of the power distribution cabinet is characterized by: The steps for implementing the monitoring and protection method for the operation safety of the power distribution cabinet as described in any one of claims 1 to 2 include: A service cross-sectional data receiving module is used to receive the service cross-sectional data of the first component of the power distribution cabinet, wherein the service cross-sectional data includes environmental humidity cross-sectional data, mechanical motion cross-sectional data and electrical parameter cross-sectional data, wherein the environmental humidity cross-sectional data represents the humidity of the environment where the power distribution cabinet is located, the mechanical motion cross-sectional data refers to the motion of the mechanical part in the power distribution cabinet, and the electrical parameter cross-sectional data refers to the current and voltage parameters of the electrical part in the power distribution cabinet; A mechanical life analysis module, used to perform mechanical life analysis based on the environmental humidity cross-sectional data and the mechanical motion cross-sectional data to obtain a mechanical life margin prediction value; An electrical life analysis module, used to perform electrical life analysis based on the environmental humidity cross-sectional data and the electrical parameter cross-sectional data to obtain an electrical life margin prediction value; A life margin prediction module, used for generating scrap identification information for the first component of the power distribution cabinet and generating a scrap signal for the associated power distribution cabinet when the mechanical life margin prediction value is less than the mechanical life margin threshold and when the electrical life margin prediction value is less than the electrical life margin threshold; The scrapping identification information sending module is used to send the scrapping identification information and the scrapping signal to the distribution cabinet operation and maintenance management terminal.

Images