Explosion-proof power distribution control system combined with safety monitoring

Abstract

The application discloses an explosion-proof power distribution control system combined with safety monitoring and relates to the related field of power distribution cabinets. When the explosion-proof power distribution cabinet is started, the first element electrical control parameter to the Nth element electrical control parameter are communicated with the explosion-proof power distribution cabinet and received. The element electrical control parameters are traversed to predict the heat generation, and the first element heat generation to the Nth element heat generation are obtained. Through the temperature sensor, the initial temperature in the cabinet and the monitoring temperature outside the cabinet are obtained. Through the cabinet temperature prediction model, the cabinet temperature time sequence curve is obtained, and the fastest triggering time length of the cabinet temperature time sequence curve not belonging to the temperature threshold interval is obtained. When the fastest triggering time length is less than or equal to the triggering time length threshold, the power distribution cabinet safety abnormal signal is generated and sent to the user end. The technical problem that the dynamic change of the internal temperature of the explosion-proof power distribution cabinet cannot be accurately predicted, the temperature overrun risk cannot be warned in advance, and the power distribution cabinet has a safety hazard is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power distribution cabinets, and particularly relates to an explosion-proof power distribution control system combined with safety monitoring. BACKGROUND

[0002] With the rapid development of industrial production, explosion-proof power distribution cabinets have been widely used in environments with explosion hazards such as petroleum, chemical industry, and mining industry. The main function of the explosion-proof power distribution cabinet is to distribute and control power while having explosion-proof performance to prevent explosion accidents caused by electrical faults. However, during the operation of the power distribution cabinet, the internal electrical components such as circuit breakers, transformers, and capacitors will generate a large amount of heat, causing the temperature inside the cabinet to gradually rise. The traditional explosion-proof power distribution usually prompts when an abnormality occurs, with a large time delay and potential safety hazards.

[0003] Therefore, in the prior art, the dynamic changes of the internal temperature of the explosion-proof power distribution cabinet cannot be accurately predicted, the temperature overrun risk cannot be warned in advance, and the technical problem of potential safety hazards of the power distribution cabinet exists. SUMMARY

[0004] The present application provides an explosion-proof power distribution control system combined with safety monitoring, which solves the technical problem that the dynamic changes of the internal temperature of the explosion-proof power distribution cabinet cannot be accurately predicted, the temperature overrun risk cannot be warned in advance, and the potential safety hazards of the power distribution cabinet exist in the prior art. By establishing a heat generation prediction model of electrical components and a temperature prediction model based on a graph neural network, the time sequence changes of the internal temperature of the power distribution cabinet are accurately predicted, the temperature overrun risk is warned in advance, and the safety and reliability of the explosion-proof power distribution cabinet are significantly improved.

[0005] The present application provides an explosion-proof power distribution control system combined with safety monitoring, which includes: a control parameter acquisition module for communicating with the explosion-proof power distribution cabinet and receiving first element electrical control parameters to Nth element electrical control parameters when the explosion-proof power distribution cabinet starts. A heat generation acquisition module is used to traverse the first element electrical control parameters to the Nth element electrical control parameters to predict heat generation and obtain first element heat generation to Nth element heat generation, wherein the first element heat generation to the Nth element heat generation represents the heat generation per unit time. A temperature monitoring module is used to obtain an initial cabinet temperature and an external monitoring temperature through a temperature sensor. A temperature curve acquisition module is used to input the initial cabinet temperature, the external monitoring temperature, the first element electrical control parameters to the Nth element electrical control parameters into a cabinet temperature prediction model to obtain a cabinet temperature time sequence curve. A trigger time acquisition module is used to obtain the fastest trigger time of the cabinet temperature time sequence curve not belonging to the temperature threshold interval. An abnormal signal sending module is used to generate a power distribution cabinet safety abnormal signal when the fastest trigger time is less than or equal to a trigger time threshold, and send the signal to a user end.

[0006] In a possible implementation, the heat generation amount obtaining module is further configured to: according to the first element type, activate a first element heat generation amount prediction model, process the first element electrical control parameter, and obtain the first element heat generation amount; and according to the Nth element type, activate an Nth element heat generation amount prediction model, process the Nth element electrical control parameter, and obtain the Nth element heat generation amount.

[0007] In a possible implementation, the heat generation amount obtaining module is further configured to: according to the electrical control parameter rated interval of the first element type, randomly assign a plurality of sets of electrical control parameter assignment results; calibrate the heat generation amount by traversing the plurality of sets of electrical control parameter assignment results, and obtain a plurality of element heat generation amount calibration data; and according to the plurality of element heat generation amount calibration data and the plurality of sets of electrical control parameter assignment results, perform supervised training to obtain the first element heat generation amount prediction model.

[0008] In a possible implementation, the heat generation amount obtaining module is further configured to: according to the plurality of sets of electrical control parameter assignment results, extract a first set of electrical control parameter assignment results; collect a set of unit time heat generation record values of the first element type under the constraint of the first set of electrical control parameter assignment results; perform mode fitting on the set of unit time heat generation record values to obtain first element heat generation amount calibration data; and add the first element heat generation amount calibration data to the plurality of element heat generation amount calibration data.

[0009] In a possible implementation, the temperature curve obtaining module is further configured to: the cabinet temperature prediction model includes a local temperature prediction model and a global temperature prediction model, wherein the local temperature prediction model includes a first element subdomain temperature prediction model to an Nth element subdomain temperature prediction model, and an input layer of the global temperature prediction model is fully connected with output layers of the first element subdomain temperature prediction model to the Nth element subdomain temperature prediction model, and a subdomain is a spatial domain constructed according to a preset radius with an element position as a center. The first element subdomain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature, and the first element electrical control parameter to obtain first element subdomain temperature time series data. The Nth element subdomain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature, and the Nth element electrical control parameter to obtain Nth element subdomain temperature time series data. The initial cabinet temperature, the external monitoring temperature, the first element subdomain temperature time series data, and the Nth element subdomain temperature time series data are input into the global temperature prediction model to obtain the cabinet temperature time series curve.

[0010] In a possible implementation, the temperature curve acquisition module is further configured to: collect initial temperature record data in the cabinet, monitored temperature record data outside the cabinet, electrical control parameter record data of the element, and sub-domain temperature time series record data of the first element type. A sub-domain temperature prediction model loss function is constructed:

[0011]

[0012] characterizes a loss value of the sub-domain temperature prediction model, characterizes an i-th time point sub-domain temperature record value of the sub-domain temperature time series record data, characterizes an i-th time point sub-domain temperature prediction value of the sub-domain temperature time series prediction data, characterizes a smoothing parameter, and m characterizes a total number of time series of the sub-domain temperature time series prediction data and the sub-domain temperature time series record data. According to the sub-domain temperature prediction model loss function, the initial temperature record data in the cabinet, the monitored temperature record data outside the cabinet, the electrical control parameter record data of the element, and the sub-domain temperature time series record data are used to train the first element sub-domain temperature prediction model.

[0013] In a possible implementation, the temperature curve acquisition module is further configured to: obtain a three-dimensional model of a cavity space in the electrical cabinet. The three-dimensional model of the cavity space in the electrical cabinet is segmented according to a preset edge length grid to obtain a grid space model of the cavity in the electrical cabinet. A graph neural network is constructed based on the grid space model of the cavity in the electrical cabinet, where a spatial topology of the graph neural network is the same as that of the grid space model of the cavity in the electrical cabinet. The graph neural network has a first node group, where the first node group is in one-to-one correspondence with grids of the grid space model of the cavity in the electrical cabinet. First element sub-domains to Nth element sub-domains are identified in the first node group to obtain a second node group, where the first node group is in one-to-one correspondence with the element sub-domains. The first node group is configured with an initial temperature input node in the cabinet and a monitored temperature input node outside the cabinet, the second node group is configured with an element sub-domain temperature time series data input node, and the graph neural network is configured with a cabinet temperature time series curve output node. Historical operation data of the electrical cabinet is collected for training to obtain the global temperature prediction model. The global temperature prediction model loss function is the same as the sub-domain temperature prediction model loss function.

[0014] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0015] ​​The application provides a safe explosion-proof power distribution control system combined with safety monitoring, comprising: a control parameter acquisition module, configured to communicate with an explosion-proof power distribution cabinet and receive first element electrical control parameters to Nth element electrical control parameters when the explosion-proof power distribution cabinet starts. A heat generation amount acquisition module is configured to traverse the first element electrical control parameters to the Nth element electrical control parameters to perform heat generation amount prediction, and obtain first element heat generation amount to Nth element heat generation amount, wherein the first element heat generation amount to the Nth element heat generation amount represents unit time heat generation amount. A temperature monitoring module is configured to obtain an initial cabinet temperature and an external cabinet monitoring temperature through a temperature sensor. A temperature curve acquisition module is configured to input the initial cabinet temperature, the external cabinet monitoring temperature, the first element electrical control parameters to the Nth element electrical control parameters into a cabinet temperature prediction model to obtain a cabinet temperature time curve. A trigger time length acquisition module is configured to obtain the fastest trigger time length of the cabinet temperature time curve not belonging to a temperature threshold interval. An abnormal signal sending module is configured to generate a power distribution cabinet safety abnormal signal and send it to a user terminal when the fastest trigger time length is less than or equal to a trigger time length threshold. The technical problem that the prior art cannot accurately predict the dynamic change of the internal temperature of the explosion-proof power distribution cabinet, cannot early warn the temperature overrun risk, and causes the explosion-proof power distribution cabinet to have a safety hazard is solved. By establishing an electrical element heat generation amount prediction model and a temperature prediction model based on a graph neural network, the time sequence change of the internal temperature of the power distribution cabinet is accurately predicted, early warning of the temperature overrun risk is realized, and the safety and reliability of the explosion-proof power distribution cabinet are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings of the embodiments of the application will be briefly introduced below. In the present application, a flowchart is used to illustrate the operations performed by the system according to the embodiments of the application. It should be understood that the foregoing or the following operations are not necessarily performed in sequence. On the contrary, various steps can be processed in reverse order or simultaneously as needed. Meanwhile, other operations can be added to these processes, or a step or several steps can be removed from these processes.

[0017] Figure 1 The structural schematic diagram of the explosion-proof power distribution control system combined with safety monitoring provided by the embodiments of the application is shown in the figure.

[0018] Figure 2 The flowchart for obtaining the first element heat generation amount to the Nth element heat generation amount in the explosion-proof power distribution control system combined with safety monitoring of the application is shown in the figure.

[0019] Legend: control parameter acquisition module 11, heat generation amount acquisition module 12, temperature monitoring module 13, temperature curve acquisition module 14, trigger time length acquisition module 15, and abnormal signal sending module 16. DETAILED DESCRIPTION

[0020] The above description is only a summary of the technical solutions of the present application. In order to make the technical means of the present application more clearly understood and implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described.

[0021] In order to make the purposes, technical solutions and advantages of the present application more clear, the following will further describe the present application in combination with the drawings, and the described embodiments should not be regarded as limitation of the present application. All other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0022] In the following description, "some embodiments" are referred to, which describe a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subset of all possible embodiments, and can be combined with each other without conflict. The term "first\second" referred to is only to distinguish similar objects, and does not represent a specific order of the objects. The terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, system, product or server including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or modules not clearly listed or inherent to these processes, systems, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as understood by those skilled in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0023] The embodiments of the present application provide an explosion-proof power distribution control system combined with safety monitoring, as shown in Figure 1 The system comprises:

[0024] A control parameter acquisition module 11 is configured to communicate with the explosion-proof power distribution cabinet and receive first element electrical control parameters to Nth element electrical control parameters when the explosion-proof power distribution cabinet is started.

[0025] A heat generation amount acquisition module 12 is configured to traverse the first element electrical control parameters to the Nth element electrical control parameters to perform heat generation amount prediction, and obtain first element heat generation amount to Nth element heat generation amount, wherein the first element heat generation amount to the Nth element heat generation amount represents unit time heat generation amount.

[0026] A temperature monitoring module 13 is configured to obtain an initial temperature in the cabinet and a monitoring temperature outside the cabinet through a temperature sensor.

[0027] Specifically, when the explosion-proof power distribution cabinet is started, the control system establishes communication with the explosion-proof power distribution cabinet and receives electrical control parameters of the first element to the Nth element. The explosion-proof power distribution cabinet is a power distribution device in a dangerous environment, and explosion-proof measures are considered in the design to prevent explosions caused by electrical faults. The electrical control parameters of the elements include operating parameters of the electrical elements, such as voltage, current, power, etc., which determine the working state and heat generation of the elements. Subsequently, heat generation is predicted by traversing the electrical control parameters of the first element to the Nth element, and the first element heat generation to the Nth element heat generation is obtained, wherein the first element heat generation to the Nth element heat generation represents the heat generation per unit time, and the heat generation per unit time is the heat generated by the element in a unit time, usually 1 hour or 1 minute. Subsequently, through the temperature sensor, the initial temperature in the cabinet and the monitoring temperature outside the cabinet are obtained, the initial temperature in the cabinet is the temperature inside the power distribution cabinet at the beginning of monitoring, and the monitoring temperature outside the cabinet is the temperature of the environment outside the power distribution cabinet.

[0028] As shown in Figure 2 The heat generation obtaining module 12 is further configured to activate a first element heat generation prediction model according to a first element model, process the electrical control parameters of the first element, and obtain the first element heat generation. Until activating an Nth element heat generation prediction model according to an Nth element model, processing the Nth element electrical control parameters, and obtaining the Nth element heat generation.

[0029] Traversing the electrical control parameters of the first element to the Nth element to predict the heat generation, and obtaining the first element heat generation to the Nth element heat generation, includes: activating a first element heat generation prediction model according to a first element model, the heat generation prediction model and the element model correspond to each other, processing the first element electrical control parameters, and obtaining the first element heat generation. Further, the heat generation of other element models is obtained, until the Nth element heat generation prediction model is activated according to the Nth element model, the Nth element electrical control parameters are processed, and the Nth element heat generation is obtained.

[0030] The heat generation obtaining module 12 is further configured to randomly assign according to the electrical control parameter rated interval of the first element model to obtain a plurality of sets of electrical control parameter assignment results. Traversing the plurality of sets of electrical control parameter assignment results to obtain a plurality of element heat generation identification data. According to the plurality of element heat generation identification data and the plurality of sets of electrical control parameter assignment results, supervised training is performed to obtain the first element heat generation prediction model.

[0031] According to the first element model, the first element heat generation prediction model is activated, the first element electrical control parameter is processed, and the first element heat generation is obtained, including: according to the electrical control parameter rated interval of the first element model, a plurality of control parameters in the electrical control parameter rated interval of the first element model are obtained by random assignment, and a plurality of sets of electrical control parameter assignment results are obtained. Subsequently, the plurality of sets of electrical control parameter assignment results are traversed to calibrate the heat generation, a plurality of element heat generation calibration data are obtained, the plurality of element heat generation calibration data correspond to the plurality of sets of electrical control parameter assignment results one by one. Finally, according to the plurality of element heat generation calibration data and the plurality of sets of electrical control parameter assignment results, supervised training is performed, the plurality of element heat generation calibration data and the plurality of sets of electrical control parameter assignment results are used as training data to supervise the training of the neural network model, until the element heat generation calibration data finally output by the model meets the preset accuracy rate, the training of the model is completed, and the first element heat generation prediction model is obtained. Further, the same way of obtaining the first element heat generation prediction model is used to obtain the second element heat generation prediction model until the Nth element heat generation prediction model, so that the element heat generation data obtained by prediction is more accurate, thereby facilitating the subsequent acquisition of the cabinet temperature curve.

[0032] The heat generation acquisition module 12 is also used to: according to the plurality of sets of electrical control parameter assignment results, a first set of electrical control parameter assignment results is extracted. With the first set of electrical control parameter assignment results as a constraint, a unit time heat generation record value set of the first element model is collected. The mode fitting is performed on the unit time heat generation record value set, and the first element heat generation calibration data is obtained. The first element heat generation calibration data is added to the plurality of element heat generation calibration data.

[0033] The heat generation of the first element is calibrated by traversing the several sets of electrical control parameter assignment results, and several element heat generation identification data are obtained, including: according to the several sets of electrical control parameter assignment results, a first set of electrical control parameter assignment results is extracted, the first set of electrical control parameter assignment results is a random set of the several sets of electrical control parameter assignment results, and each set of electrical control parameter assignment results contains all control parameters of the first element. Then, with the first set of electrical control parameter assignment results as a constraint, a unit time heat generation record value set of the first element model is collected, that is, in the historical first element operation data or operation experiment data, the first set of electrical control parameter assignment results is taken as a control constraint to collect the unit time heat generation record value set of the first set of electrical control parameter assignment results, and the unit time can be set according to actual conditions. Further, the mode of the unit time heat generation record value set is fitted to obtain the mode in the unit time heat generation record value set, and the first element heat generation identification data are obtained. Finally, the first element heat generation identification data are added to the several element heat generation identification data, and the same way is used to obtain the heat generation identification data corresponding to the several sets of electrical control parameter assignment results and add them.

[0034] The temperature curve acquisition module 14 is used to input the initial temperature in the cabinet, the monitoring temperature outside the cabinet, the first element electrical control parameter to the Nth element electrical control parameter into the cabinet temperature prediction model to obtain a cabinet temperature time curve. The trigger time length acquisition module 15 is used to obtain the fastest trigger time length of the cabinet temperature time curve not belonging to the temperature threshold interval. The abnormal signal sending module 16 is used to generate a power distribution cabinet safety abnormal signal when the fastest trigger time length is less than or equal to the trigger time length threshold, and send the signal to the user end.

[0035] The obtained initial temperature in the cabinet, the monitoring temperature outside the cabinet, the first element electrical control parameter to the Nth element electrical control parameter are input into the cabinet temperature prediction model, and temperature prediction is performed through the cabinet temperature prediction model to obtain a cabinet temperature time curve. Further, the fastest trigger time length of the cabinet temperature time curve not belonging to the temperature threshold interval is obtained, the temperature threshold interval is the temperature range for safe operation of the power distribution cabinet, and the upper and lower limit temperatures are determined by the equipment specifications or safety standards. The fastest trigger time length is the shortest time required for the temperature to exceed the safety threshold. When the fastest trigger time length is less than or equal to the trigger time length threshold, a power distribution cabinet safety abnormal signal is generated and sent to the user end, and the trigger time length threshold is a preset time threshold for judging the urgency of temperature overrun risk. When the fastest trigger time length is less than or equal to the trigger time length threshold, the temperature of the power distribution cabinet obviously exceeds the temperature safety threshold in a short time, and emergency treatment is required. When the fastest trigger time length is greater than the trigger time length threshold, the temperature rises slowly, and the system abnormality is small, and no abnormal signal is sent.

[0036] The temperature curve acquisition module 14 is further configured to: the cabinet temperature prediction model comprises a local temperature prediction model and a global temperature prediction model, wherein the local temperature prediction model comprises a first element sub-domain temperature prediction model to an Nth element sub-domain temperature prediction model, the input layer of the global temperature prediction model is fully connected with the output layer of the first element sub-domain temperature prediction model to the Nth element sub-domain temperature prediction model, and a sub-domain is a spatial domain constructed according to a preset radius with an element position as a center. The first element sub-domain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature and the first element electrical control parameter to obtain first element sub-domain temperature time series data. The Nth element sub-domain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature and the Nth element electrical control parameter to obtain Nth element sub-domain temperature time series data. The initial cabinet temperature, the external monitoring temperature, the first element sub-domain temperature time series data to the Nth element sub-domain temperature time series data are input into the global temperature prediction model to obtain the cabinet temperature time series curve.

[0037] The cabinet temperature prediction model comprises a local temperature prediction model and a global temperature prediction model, wherein the local temperature prediction model comprises a first element sub-domain temperature prediction model to an Nth element sub-domain temperature prediction model, each element sub-domain temperature prediction model corresponds to an element model. The input layer of the global temperature prediction model is fully connected with the output layer of the first element sub-domain temperature prediction model to the Nth element sub-domain temperature prediction model, and a sub-domain is a spatial domain constructed according to a preset radius with an element position as a center. Further, the first element sub-domain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature and the first element electrical control parameter to obtain first element sub-domain temperature time series data. The Nth element sub-domain temperature prediction model is used to process the initial cabinet temperature, the external monitoring temperature and the Nth element electrical control parameter to obtain Nth element sub-domain temperature time series data. N is a positive integer, and N is equal to the number of element models. The initial cabinet temperature, the external monitoring temperature, the first element sub-domain temperature time series data to the Nth element sub-domain temperature time series data are input into the global temperature prediction model to obtain the cabinet temperature time series curve.

[0038] The temperature curve acquisition module 14 is further configured to: collect initial cabinet temperature record data, external monitoring temperature record data, element electrical control parameter record data and sub-domain temperature time series record data of a first element model.

[0039] a sub-domain temperature prediction model loss function is constructed:

[0040] ,

[0041] wherein, characterizes a sub-domain temperature prediction model loss value, characterizes an i-th time point sub-domain temperature record value of a sub-domain temperature time series record data, characterizes an i-th time point sub-domain temperature prediction value of a sub-domain temperature time series prediction data, characterizes a smoothing parameter, and m characterizes a total number of time series of the sub-domain temperature time series prediction data and the sub-domain temperature time series record data. According to the sub-domain temperature prediction model loss function, the cabinet initial temperature record data, the cabinet external monitoring temperature record data, the component electrical control parameter record data and the sub-domain temperature time series record data are called to train the first component sub-domain temperature prediction model.

[0042] The first component sub-domain temperature prediction model training step includes: collecting cabinet initial temperature record data, cabinet external monitoring temperature record data, component electrical control parameter record data and sub-domain temperature time series record data recorded in the historical operation process of the first component model, wherein the sub-domain temperature time series record data is the average temperature data in the region recorded in time series in the spatial domain constructed with the component position as the center and according to a preset radius. A sub-domain temperature prediction model loss function is constructed:

[0043] ,

[0044] wherein, characterizes a sub-domain temperature prediction model loss value, characterizes an i-th time point sub-domain temperature record value of a sub-domain temperature time series record data, characterizes an i-th time point sub-domain temperature prediction value of a sub-domain temperature time series prediction data, characterizes a smoothing parameter, and m characterizes a total number of time series of the sub-domain temperature time series prediction data and the sub-domain temperature time series record data, The smoothing parameter is usually 1, 0.1, 0.01 and 0.001, and finally the optimal parameter value is selected as the target smoothing parameter.

[0045] According to the sub-domain temperature prediction model loss function, the cabinet initial temperature record data, the cabinet outside monitoring temperature record data, the element electrical control parameter record data and the sub-domain temperature time series record data are called as training data, the neural network model is supervised trained, the sub-domain temperature time series record data outputted after any group of data is inputted into the model at the end of model training is obtained, and the sub-domain temperature prediction model loss value calculation is performed through the sub-domain temperature prediction model loss function. When the sub-domain temperature prediction model loss value is less than the preset loss threshold value, a plurality of data in the training data is randomly selected as verification data, the results outputted by the verification data are obtained, and the loss value calculation is performed through the sub-domain temperature prediction model loss function, until more than 98% of the calculation results in the loss value calculation results are less than the preset loss threshold value, the training of the model is completed, and the first element sub-domain temperature prediction model is obtained.

[0046] The temperature curve acquisition module 14 is further configured to obtain a three-dimensional model of the inner cavity space of the electrical cabinet. The three-dimensional model of the inner cavity space of the electrical cabinet is segmented according to a preset edge length grid to obtain a grid space model of the inner cavity of the electrical cabinet. A graph neural network is constructed based on the grid space model of the inner cavity of the electrical cabinet, wherein the spatial topology of the graph neural network is the same as that of the grid space model of the inner cavity of the electrical cabinet. The graph neural network has a first node group, wherein the first node group corresponds to the grids of the grid space model of the inner cavity of the electrical cabinet one by one. The first node group is identified by the first element sub-domain to the Nth element sub-domain to obtain a second node group, wherein the second node group corresponds to the element sub-domains one by one. The first node group is configured with a cabinet initial temperature input node and a cabinet outside monitoring temperature input node, the second node group is configured with an element sub-domain temperature time series data input node, and the graph neural network is configured with a cabinet temperature time series curve output node. Historical operation data of the electrical cabinet is collected for training to obtain the global temperature prediction model.

[0047] A three-dimensional model of the inner space of the electrical cabinet is obtained, which is a three-dimensional digital representation of the inner space of the electrical cabinet, including the physical size, shape and element layout in the cabinet. Then, the three-dimensional model of the inner space of the electrical cabinet is segmented according to a preset grid length, and the continuous three-dimensional space is segmented into a plurality of small, discrete grid units according to the preset grid length, to obtain an electrical cabinet inner space grid model. The specific preset grid length can be set according to the actual accuracy preference, and the higher the accuracy, the smaller the corresponding grid length. Further, a graph neural network with the same topological structure as the electrical cabinet inner space grid model is constructed based on the electrical cabinet inner space grid model, wherein the spatial topology of the graph neural network is the same as that of the electrical cabinet inner space grid model. The graph neural network has a first node group, wherein the first node group corresponds one-to-one to the grid of the electrical cabinet inner space grid model. The first element sub-domain to the Nth element sub-domain are identified in the first node group, i.e., each element sub-domain is identified in the first node group, to obtain a second node group, wherein the second node group region corresponds one-to-one to the element sub-domain. Finally, the first node group is configured with an initial temperature input node in the cabinet and a monitored temperature input node outside the cabinet, the second node group is configured with an element sub-domain temperature time series data input node, and the graph neural network is configured with a cabinet temperature time series curve output node.

[0048] Further, the initial temperature data of the first node group and the corresponding cabinet outside monitoring temperature and the sub-domain temperature time series data of each region in the second node group and the corresponding cabinet temperature time series data in the collected electrical cabinet historical operation data are collected as training data. The graph neural network is trained until the temperature time series data output by the graph neural network is verified by the same verification method as the sub-domain temperature prediction model for the global temperature prediction model loss function verification, until more than 98% of the calculation results in the loss value calculation result are less than the preset loss threshold, the training of the model is completed, and the global temperature prediction model is obtained. The global temperature prediction model loss function is the same as the sub-domain temperature prediction model loss function. The global temperature prediction model also has a cabinet temperature time series curve output node configured therein, which is used to draw a cabinet temperature time series curve according to the cabinet temperature time series data output by the model.

[0049] The embodiment of the application obtains the first element heat quantity to the Nth element heat quantity by executing the steps of communicating with the explosion-proof power distribution cabinet and receiving the first element electrical control parameter to the Nth element electrical control parameter when the explosion-proof power distribution cabinet starts. The first element heat quantity to the Nth element heat quantity is obtained by traversing the first element electrical control parameter to the Nth element electrical control parameter to perform heat quantity prediction, wherein the first element heat quantity to the Nth element heat quantity represents the heat quantity per unit time. The initial cabinet temperature and the external cabinet monitoring temperature are obtained by the temperature sensor. The initial cabinet temperature, the external cabinet monitoring temperature, the first element electrical control parameter to the Nth element electrical control parameter are input into the cabinet temperature prediction model to obtain the cabinet temperature time sequence curve. The fastest triggering time length that does not belong to the temperature threshold interval is obtained from the cabinet temperature time sequence curve. When the fastest triggering time length is less than or equal to the triggering time length threshold, the power distribution cabinet safety abnormal signal is generated and sent to the user end. The technical problem that the dynamic change of the internal temperature of the explosion-proof power distribution cabinet cannot be accurately predicted, the temperature overrun risk cannot be warned in advance, and the safety hazard of the power distribution cabinet exists in the prior art is solved. By establishing the heat quantity prediction model of the electrical element and the temperature prediction model based on the graph neural network, the time sequence change of the internal temperature of the power distribution cabinet is accurately predicted, the temperature overrun risk is warned in advance, and the safety and reliability of the explosion-proof power distribution cabinet are significantly improved.

[0050] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application. In some cases, the actions or steps described in the present application can be executed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

Claims

1. An explosion-proof power distribution control system combined with safety monitoring, characterized in that, include: The control parameter acquisition module is used to communicate with the explosion-proof distribution cabinet and receive the electrical control parameters of the first element up to the electrical control parameters of the Nth element when the explosion-proof distribution cabinet is started. The heat generation acquisition module is used to traverse the electrical control parameters of the first element up to the electrical control parameters of the Nth element to predict the heat generation, and obtain the heat generation of the first element up to the Nth element, wherein the heat generation of the first element up to the Nth element represents the heat generation per unit time. The temperature monitoring module is used to obtain the initial temperature inside the cabinet and the monitored temperature outside the cabinet through a temperature sensor; The temperature curve acquisition module is used to input the initial temperature inside the cabinet, the monitored temperature outside the cabinet, the electrical control parameters of the first component up to the electrical control parameters of the Nth component into the cabinet temperature prediction model to obtain the time series curve of the cabinet temperature. The trigger duration acquisition module is used to obtain the fastest trigger duration for which the temperature time sequence curve inside the cabinet does not belong to the temperature threshold range. An abnormal signal sending module is used to generate a power distribution cabinet safety abnormal signal and send it to the user terminal when the fastest trigger duration is less than or equal to the trigger duration threshold. The heat generation acquisition module is also used for: Based on the model number of the first component, activate the heat generation prediction model of the first component, process the electrical control parameters of the first component, and obtain the heat generation of the first component; Until the heat generation prediction model of the Nth component is activated according to the model number of the Nth component, the electrical control parameters of the Nth component are processed to obtain the heat generation of the Nth component; The temperature curve acquisition module is also used for: The cabinet temperature prediction model includes a local temperature prediction model and a global temperature prediction model. The local temperature prediction model includes a first element subdomain temperature prediction model up to the Nth element subdomain temperature prediction model. The input layer of the global temperature prediction model is fully connected to the output layer of the first element subdomain temperature prediction model up to the Nth element subdomain temperature prediction model. The subdomain is a spatial domain constructed with the element position as the center according to a preset radius. The initial temperature inside the cabinet, the monitored temperature outside the cabinet, and the electrical control parameters of the first element are processed by the temperature prediction model of the first element subdomain to obtain the time series data of the temperature of the first element subdomain. The initial temperature inside the cabinet, the monitored temperature outside the cabinet, and the electrical control parameters of the Nth element are processed by the temperature prediction model of the Nth element subdomain to obtain the time series data of the temperature of the Nth element subdomain. The initial temperature inside the cabinet, the monitored temperature outside the cabinet, the time series data of the temperature of the first element subdomain up to the time series data of the temperature of the Nth element subdomain are input into the global temperature prediction model to obtain the time series curve of the temperature inside the cabinet.

2. The explosion-proof power distribution control system combined with safety monitoring as described in claim 1, characterized in that, The heat generation acquisition module is also used for: Random values ​​are assigned to the electrical control parameters within the rated range of the first component model to obtain several sets of electrical control parameter assignment results; The heat generation is calibrated by iterating through the several sets of electrical control parameter assignment results, and heat generation identification data of several components is obtained; Supervised training is performed based on the heat generation identification data of the aforementioned components and the assignment results of the aforementioned sets of electrical control parameters to obtain the heat generation prediction model of the first component.

3. The explosion-proof power distribution control system combined with safety monitoring as described in claim 2, characterized in that, The heat generation acquisition module is also used for: Based on the several sets of electrical control parameter assignment results, extract the first set of electrical control parameter assignment results; Using the first set of electrical control parameter assignment results as constraints, collect a set of heat generation records per unit time for the first component model; The mode of the set of heat output records per unit time is fitted to obtain the heat output identification data of the first element; Add the heat output identification data of the first component to the heat output identification data of the plurality of components.

4. The explosion-proof power distribution control system combined with safety monitoring as described in claim 1, characterized in that, The temperature curve acquisition module is also used for: Collect initial temperature records inside the cabinet, external temperature records, electrical control parameter records, and sub-domain temperature time sequence records for the first component model; Constructing the loss function for the subdomain temperature prediction model: ， in, Characterizes the loss value of the subdomain temperature prediction model. The temperature record value of the subdomain at time i, representing the temperature time series data of the subdomain. The predicted temperature value of the subdomain at time i, representing the time-series temperature prediction data of the subdomain. The smoothing parameter is represented by m, which represents the total number of time series predicted data and recorded data of subdomain temperature time series. Based on the loss function of the subdomain temperature prediction model, the initial temperature record data inside the cabinet, the temperature record data monitored outside the cabinet, the electrical control parameter record data of the component, and the time-series temperature record data of the subdomain are retrieved to train the first component subdomain temperature prediction model.

5. The explosion-proof power distribution control system combined with safety monitoring as described in claim 4, characterized in that, The temperature curve acquisition module is also used for: Obtain a three-dimensional model of the internal space of the electrical cabinet; The three-dimensional model of the electrical cabinet's internal cavity space is divided according to a preset side length grid to obtain a grid space model of the electrical cabinet's internal cavity space. A graph neural network is constructed based on the internal grid space model of the electrical cabinet, wherein the spatial topology of the graph neural network is the same as that of the internal grid space model of the electrical cabinet. The graph neural network has a first node group, wherein the first node group corresponds one-to-one with the grid of the electrical cabinet cavity grid space model; The first element subdomain up to the Nth element subdomain are identified in the first node group to obtain the second node group, wherein the first node group corresponds one-to-one with the element subdomain; Configure the first node group with an initial temperature input node inside the cabinet and an external temperature monitoring input node, configure the second node group with a component subdomain temperature time series data input node, configure the graph neural network with an internal temperature time series curve output node, collect historical operating data of the electrical cabinet for training, and obtain the global temperature prediction model.

6. The explosion-proof power distribution control system combined with safety monitoring as described in claim 5, characterized in that, The loss function of the global temperature prediction model is the same as that of the subdomain temperature prediction model.

Images