Switchgear state monitoring method and system based on artificial intelligence

Abstract

The embodiment of the application provides a switch cabinet state monitoring method and system based on artificial intelligence, by acquiring a state update activity combination and a classification index combination of a target switch cabinet control system, and distinguishing between a leading state activity and a reactive state activity, the activity session category of each state update activity combination is determined, which can not only help operation and maintenance personnel to more effectively manage and monitor the switch cabinet control task, but also timely report the exception when the control unit instance is abnormal, so as to avoid or reduce the possible loss. In addition, the abnormality detection mechanism based on the activity session category can more accurately locate the abnormality, whether it is a time domain or an operation domain, and the abnormality can be accurately captured and reported to the switch cabinet system operation and maintenance center. In this way, the operation and maintenance personnel can quickly and accurately find the problem and take necessary repair measures to ensure the stable operation of the switch cabinet control system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a switch cabinet state monitoring method and system based on artificial intelligence. BACKGROUND

[0002] In the traditional switch cabinet control system, various state update activities usually need to be monitored and managed manually. For example, the operation and maintenance personnel need to regularly check the state of the switch cabinet and analyze various state update activities to ensure the normal operation of the switch cabinet control system. However, this method has some significant drawbacks.

[0003] Firstly, manual monitoring and management is a time-consuming and inefficient process. In a large or complex switch cabinet control system, there may be a large number of state update activities, and manual monitoring and management will occupy a large amount of human resources.

[0004] Secondly, the accuracy of manual monitoring and management is affected by human factors. For example, the operation and maintenance personnel may have difficulty in effectively managing proactive state activities and reactive state activities, or understanding and handling some complex state update activities are not accurate enough. Therefore, how to improve the intelligent level of the switch cabinet control system, realize more efficient and accurate state update activity monitoring and management, and effective anomaly detection, is a problem to be solved in the current switch cabinet control technology field. SUMMARY

[0005] Therefore, the purpose of the present application is to provide a switch cabinet state monitoring method and system based on artificial intelligence.

[0006] According to a first aspect of the present application, a switch cabinet state monitoring method based on artificial intelligence is provided, applied to a state monitoring system, the method comprising:

[0007] obtaining X state update activity combinations and X classification index combinations of a target switch cabinet control system, X being a positive integer greater than 1, each state update activity combination in the X state update activity combinations comprising a proactive state activity and a reactive state activity of a pending activity session category, the X classification index combinations comprising an activity classification index of each proactive state activity and each reactive state activity in the X state update activity combinations, each state update activity combination comprising one proactive state activity with an active trigger state and one reactive state activity with a passive trigger state, the proactive state activity comprising one proactive element in a target switch cabinet control task, the proactive element comprising a preset operation rule or one instruction sequence, the reactive state activity comprising one reactive element in the target switch cabinet control task, the reactive element comprising a feedback mechanism or an adaptive strategy;

[0008] determine X activity session categories of the X state update activity combinations based on X dominant state description features, X first index category description features, X reactive state description features, and X second index category description features, the X dominant state description features being description vectors of X dominant state activities in the X state update activity combinations, the X first index category description features being index category features of the X dominant state activities, the X reactive state description features being description vectors of X reactive state activities in the X state update activity combinations, the X first index category description features being index category features of the X dominant state activities, and the X second index category description features being index category features of the X reactive state activities;

[0009] when a control unit instance in a target node represented by any one of the X activity session categories is abnormal, report target instance abnormality information to a switch cabinet system operation and maintenance center, the control unit instance in the target node being abnormal includes a time domain and / or an operation domain responded by the control unit instance being abnormal, and the target instance abnormality information reflecting the control unit instance in the target node being abnormal.

[0010] In a possible implementation of the first aspect, the determining the X activity session categories of the X state update activity combinations based on the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features includes:

[0011] obtaining description vectors of X dominant state activities in the X state update activity combinations to generate X dominant state description features, obtaining description vectors of X reactive state activities in the X state update activity combinations to generate X reactive state description features, obtaining index category features of activity classification indexes of the X dominant state activities in the X classification index combinations to generate X first index category description features, and obtaining index category features of activity classification indexes of the X reactive state activities in the X classification index combinations to generate X second index category description features;

[0012] fusing the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features to generate X target feature representations corresponding to the X state update activity combinations;

[0013] Based on the X target feature representations, X activity session categories of the X state update activity combinations are determined, each of the X activity session categories reflecting a session relationship between the dominant state activity and the reactive state activity in a corresponding one of the X state update activity combinations.

[0014] In a possible implementation of the first aspect, the obtaining of the description vector of the rth dominant state activity in the X dominant state activities and the generation of the rth dominant state description feature include:

[0015] The description vector of the rth dominant state activity in the X dominant state activities is obtained and the rth dominant state description feature is generated based on the following operations, r being a positive integer not less than 1 and not greater than X:

[0016] When the rth dominant state activity includes Yr state behaviors, a behavior feature representation of each of the Yr state behaviors is obtained and Yr behavior feature representations are generated, Yr being a positive integer;

[0017] The Yr behavior feature representations are balanced to generate the rth dominant state description feature, or the Yr behavior feature representations are fused by weights to generate the rth dominant state description feature.

[0018] In a possible implementation of the first aspect, the obtaining of the description vector of the rth reactive state activity in the X reactive state activities and the generation of the rth reactive state description feature include:

[0019] The description vector of the rth reactive state activity in the X reactive state activities is obtained and the rth reactive state description feature is generated based on the following operations, r being a positive integer not less than 1 and not greater than X:

[0020] When the rth reactive state activity includes Zr state behaviors, a behavior feature representation of each of the Zr state behaviors is obtained and Zr behavior feature representations are generated, Zr being a positive integer;

[0021] The Zr behavior feature representations are balanced to generate the rth reactive state description feature, or the Zr behavior feature representations are fused by weights to generate the rth reactive state description feature.

[0022] In a possible implementation of the first aspect, the obtaining of the index category feature of the activity classification index of the X dominant state activities in the X classification index combinations and the generation of X first index category description features include:

[0023] The index category feature of the activity classification index of the rth dominant state activity in the X dominant state activities is obtained based on the following operation, and an rth first index category description feature is generated, where r is a positive integer not less than 1 and not greater than X:

[0024] When the activity classification index of the rth dominant state activity includes an rth index description field, a field feature representation of the rth index description field is obtained, and the rth first index category description feature is the field feature representation of the rth index description field.

[0025] In a possible implementation of the first aspect, the index category feature of the activity classification index of the X reactive state activities in the X classification index combination is obtained, and X second index category description features are generated, including:

[0026] The index category feature of the activity classification index of the rth reactive state activity in the X reactive state activities is obtained based on the following operation, and an rth second index category description feature is generated, where r is a positive integer not less than 1 and not greater than X:

[0027] When the activity classification index of the rth reactive state activity includes an rth index description field, a field feature representation of the rth index description field is obtained, and the rth second index category description feature is the field feature representation of the rth index description field.

[0028] In a possible implementation of the first aspect, the X target feature representations corresponding to the X state update activity combinations are generated by fusing the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features, including:

[0029] The rth target feature representation corresponding to the rth state update activity combination in the X state update activity combinations is obtained based on the following operation, where the rth state update activity combination includes an rth dominant state activity and an rth reactive state activity, and r is a positive integer not less than 1 and not greater than X:

[0030] The task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are fused to generate the rth target feature representation, the task trigger description vector reflects a trigger node of a target switch cabinet control task, the state update activities in the X state update activity combination are state update activities parsed from the target switch cabinet control task, the rth dominant state description feature is a description vector of the rth dominant state activity, the rth first index category description feature is an index category feature of an activity classification index of the rth dominant state activity, the rth reactive state description feature is a description vector of the rth reactive state activity, and the rth second index category description feature is an index category feature of an activity classification index of the rth reactive state activity.

[0031] When the dimensions of the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are all 1*K, the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are fused to generate the rth target feature representation with a dimension of 5*K, and K is a positive integer greater than 1.

[0032] In a possible implementation of the first aspect, the determination of the X activity session categories of the X state update activity combinations based on the X target feature representations includes:

[0033] The rth activity session category of the rth state update activity combination in the X state update activity combinations is determined based on the rth target feature representation in the X target feature representations, and r is a positive integer not less than 1 and not greater than X.

[0034] The rth target feature representation is loaded into an AI neural network to generate M session category classification data corresponding to M activity session categories, the M session category classification data reflect the possibility that the rth activity session category is each candidate session category in the M set activity session categories, and M is a positive integer greater than 1.

[0035] The rth activity session category is determined to be equal to a target activity session category in the M set activity session categories, and the possibility of the rth activity session category being the target activity session category is the largest in the M session category classification data.

[0036] In a possible implementation of the first aspect, the determining of the X activity session categories of the X state update activity combinations based on the X target feature representations comprises:

[0037] loading the X target feature representations into the AI neural network round by round to generate a sequence of session category classification data corresponding to each round target feature representation, each round target feature representation comprising J target feature representations of the X target feature representations, and each sequence of session category classification data corresponding to each round target feature representation comprising J groups of session category classification data, each group of session category classification data comprising M session category classification data, and the M session category classification data covered by the s-th group of session category classification data reflecting the possibility that the s-th activity session category corresponding to the s-th target feature representation is each candidate session category of the M set activity session categories, the s-th group of session category classification data corresponding to the s-th target feature representation of the J target feature representations, M being a positive integer greater than 1, J being a positive integer not less than 2 and less than X, and s being a positive integer not less than 1 and not greater than J;

[0038] determining the s-th activity session category of the J activity session categories based on the s-th group of session category classification data in the sequence of session category classification data by using the following operations:

[0039] determining the s-th activity session category to be a target activity session category of the M set activity session categories, the s-th activity session category being the target activity session category having the maximum possibility among the M session category classification data covered by the s-th group of session category classification data.

[0040] For example, in a possible implementation of the first aspect, the obtaining of the X state update activity combinations and the X classification index combinations of the target switch cabinet control system comprises:

[0041] analyzing H state update activities and H activity classification indexes having unique mapping relationships in the target switch cabinet control task, each of the H activity classification indexes comprising an activity classification index of each of the H state update activities, and each of the H state update activities comprising an element in the target switch cabinet control task;

[0042] two state update activities in the one group of two state update activities are a dominant state activity and a reactive state activity;

[0043] determining, in the H activity classification indexes, an activity classification index of the dominant state activity and the reactive state activity covered in each of the X state update activity combinations, and generating X classification index combinations;

[0044] wherein the conversation relationship between the dominant state activity and the reactive state activity represented by each of the X activity conversation categories is one of M conversation relationships, M is a positive integer greater than 1, and the M conversation relationships include M-1 set conversation relationships and no conversation relationship.

[0045] According to a second aspect of the present application, a state monitoring system is provided, which comprises a processor and a readable storage medium, and the readable storage medium stores a program which, when executed by the processor, implements the aforementioned switch cabinet state monitoring method based on artificial intelligence.

[0046] According to a third aspect of the present application, a computer readable storage medium is provided, which stores computer executable instructions, and when the computer executable instructions are executed, the aforementioned switch cabinet state monitoring method based on artificial intelligence is implemented.

[0047] According to any one of the above aspects, in the present application, by obtaining the state update activity combination and its classification index combination of the target switch cabinet control system, and distinguishing the dominant state activity and the reactive state activity, the activity conversation category of each state update activity combination is determined, which not only helps the operation and maintenance personnel to more effectively manage and monitor the switch cabinet control task, but also timely reports the exception when the control unit instance is abnormal, thereby avoiding or reducing the possible loss. In addition, the exception detection mechanism based on the activity conversation category can more accurately locate the exception, whether it is a time domain or an operation domain exception, which can be accurately captured and reported to the switch cabinet system operation and maintenance center. In this way, the operation and maintenance personnel can quickly and accurately find the problem and take necessary repair measures to ensure the stable operation of the switch cabinet control system. That is, the present application significantly improves the intelligent degree and operation efficiency of the switch cabinet control system, and also provides a powerful tool for realizing accurate exception detection and rapid fault repair. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0049] Figure 1 The flowchart of the switch cabinet state monitoring method based on artificial intelligence provided by the embodiments of the present application is shown.

[0050] Figure 2 The component structure schematic diagram of the state monitoring system for implementing the above-mentioned switch cabinet state monitoring method based on artificial intelligence provided by the embodiments of the present application is shown. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. All other embodiments obtained by those skilled in the art without creative labor on the basis of the embodiments in the present application should belong to the scope of protection of the present application.

[0052] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0053] Figure 1 The flowchart of the switch cabinet state monitoring method based on artificial intelligence provided by the embodiments of the present application is shown. It should be understood that in other embodiments, the order of some steps of the switch cabinet state monitoring method based on artificial intelligence of the present embodiment can be exchanged according to actual needs, or some steps can be omitted or deleted. The detailed steps of the switch cabinet state monitoring method based on artificial intelligence are described as follows.

[0054] Step S110, obtaining X state update activity combinations and X classification index combinations of the target switch cabinet control system.

[0055] In this embodiment, X is a positive integer greater than 1, each of the X state update activity combinations includes a proactive state activity and a reactive state activity of the pending activity session category, the X classification index combinations include an activity classification index of each proactive state activity and each reactive state activity in the X state update activity combinations, each state update activity combination includes one proactive state activity with an active trigger state and one reactive state activity with a passive trigger state, the proactive state activity includes one proactive element in the target switch cabinet control task, the proactive element includes a preset operation rule or an instruction sequence, and the reactive state activity includes one reactive element in the target switch cabinet control task, the reactive element includes a feedback mechanism or an adaptive strategy.

[0056] For example, assuming that in a switch cabinet control system, X = 3, three state update activity combinations need to be processed. Each state update activity combination contains one proactive state activity and one reactive state activity.

[0057] For example, the first state update activity combination may involve load scheduling. The proactive state activity includes power grid load balancing operations performed according to a preset operation rule or an instruction sequence, such as a power scheduling plan arranged in advance according to a prediction algorithm; the reactive state activity may be real-time power adjustment according to a feedback mechanism or an adaptive strategy when the actual power grid load is significantly different from the prediction. The proactive state activity and the reactive state activity will have their own classification indexes to identify the specific categories or states of the proactive state activity and the reactive state activity.

[0058] For example, in a power switch cabinet control system, there may be a target task of "maintaining power grid load balance during peak hours". In order to complete this task, a series of state update activities may be required.

[0059] The proactive state activity includes performing the task according to a preset operation rule or an instruction sequence. In this example, the proactive element may be "setting the opening and closing plan of the switch cabinet in advance according to the power demand prediction and the grid capacity". This preset operation rule or instruction sequence will determine which switch cabinets need to be opened and which need to be closed during peak hours, and when to perform these operations.

[0060] The reactive state activity involves adjusting the system based on feedback mechanisms or adaptive strategies when there are deviations between the actual situation and the pre-planned schedule. In this example, the reactive element could be "if the actual power demand exceeds the forecast, or if a certain switch cabinet fails, the system needs to immediately adjust the status of other switch cabinets to maintain the balance of the power grid load." This feedback mechanism or adaptive strategy will ensure that the system can flexibly respond to changes in the actual situation and always maintain the stable operation of the power grid.

[0061] During this process, the proactive state activity and the reactive state activity work together to achieve the target task of the switch cabinet control system.

[0062] In step S120, X activity session categories of the X state update activity combinations are determined based on the X proactive state description features, the X first index category description features, the X reactive state description features, and the X second index category description features.

[0063] In this embodiment, the X proactive state description features are description vectors of X proactive state activities in the X state update activity combinations, the X first index category description features are index category features of the X proactive state activities, the X reactive state description features are description vectors of X reactive state activities in the X state update activity combinations, the X first index category description features are index category features of the X proactive state activities, and the X second index category description features are index category features of the X reactive state activities.

[0064] For example, when determining the activity session category, the activity session category of each state update activity combination needs to be determined based on the X proactive state description features, the X first index category description features, the X reactive state description features, and the X second index category description features.

[0065] For example, for the state update activity combination of load scheduling, the following features need to be considered:

[0066] Proactive state description features: This is a description of proactive state activities. Each proactive state activity has a description vector associated with it, which contains various information describing the proactive state activity. For example, for the aforementioned "based on power demand forecast and grid capacity, set the opening and closing plan of the switch cabinet in advance" proactive state activity, the description features may include the preset operation rules (how to set the opening and closing plan of the switch cabinet), the instruction sequence (the specific opening and closing plan), the execution time, etc.

[0067] The first index category description feature is a classification of the dominant state activity. It describes the category or state to which the dominant state activity belongs. For example, the dominant state activity can be classified as "load scheduling", "device inspection", etc. of different types, and can also be marked as "in progress", "completed", etc. of different execution states.

[0068] The reactive state description feature is a description of the reactive state activity. Each reactive state activity has a description vector associated with it, which contains various information describing the reactive state activity. For example, for the aforementioned reactive state activity "if the actual power demand exceeds the prediction, or a certain switch cabinet fails, the state of other switch cabinets needs to be adjusted immediately to maintain the load balance of the power grid", the description feature can include feedback mechanism (how to adjust the state of the switch cabinet according to the actual situation), adaptive strategy (when to start the adaptive strategy, how to adjust), execution result, etc.

[0069] The second index category description feature is a classification of the reactive state activity. It describes the category or state to which the reactive state activity belongs. For example, the reactive state activity can be classified as "load adjustment", "fault handling", etc. of different types, and can also be marked as "in progress", "completed", etc. of different execution states.

[0070] Through these features, each state update activity combination can be more accurately understood and managed for subsequent monitoring and management.

[0071] In step S130, when the control unit instance in the target node is abnormal based on any one of the X activity session categories, the target instance abnormal information is reported to the switch cabinet system operation and maintenance center.

[0072] In this embodiment, the control unit instance in the target node is abnormal, including the time domain and / or operation domain of the control unit instance response being abnormal, and the target instance abnormal information reflects the abnormality of the control unit instance in the target node.

[0073] For example, in step S120, the activity session category is determined by the dominant state description feature, the first index category description feature, the reactive state description feature, and the second index category description feature.

[0074] The dominant state description feature and the first index category description feature are used to describe and classify the dominant state activity, and the reactive state description feature and the second index category description feature are used to describe and classify the reactive state activity. Each state update activity combination (including a dominant state activity and a reactive state activity) will be assigned to a specific activity session category according to these features.

[0075] For example, if a target feature representation of a state update activity combination indicates that its dominant state activity is ongoing and the reactive state activity has been completed, it can be assigned to an activity session category named "normal operation". If the dominant state activity fails and needs to be handled immediately, it can be assigned to an activity session category named "abnormal repair". If an emergency situation is detected, such as the power grid load exceeding a threshold, and the reactive state activity fails to respond in time, it can be assigned to an activity session category named "emergency handling".

[0076] When a control unit instance in a target node is represented based on any one of the X activity session categories (e.g., an "abnormal repair" or "emergency handling" activity session category), it means that if the control unit instance fails to perform a state update activity combination corresponding to the activity session category, for example, it fails to schedule power according to the preset operation rules or cannot correctly respond to real-time changes in the power grid load, the failure is considered as an abnormality under the activity session category.

[0077] For example, a target node can refer to a specific switch cabinet or a group of switch cabinets, and a control unit instance is a device or system responsible for performing specific tasks (such as load scheduling, power grid monitoring, etc.). Time domain abnormalities are usually related to time. For example, if a control unit instance needs to complete a specific state update activity within a specified time, but the actual execution time exceeds the expectation, it can be considered to have a time domain abnormality. Another example is that the reaction time of the control unit instance becomes significantly slower, or the execution frequency of its state update activity changes abnormally. Operation domain abnormalities are usually related to the function or operation of the control unit instance. For example, if a control unit instance responsible for load scheduling fails to correctly allocate power, or a control unit instance responsible for power grid monitoring cannot accurately read the power grid state, it can be considered to have an operation domain abnormality. The target instance abnormality information reflects the abnormality of the control unit instance in the target node, which means that when the control unit instance has any of the above abnormalities, the corresponding abnormality information is generated. These abnormality information can include the time of abnormality occurrence, the specific abnormality type (time domain or operation domain), the state information of the control unit instance, the state update activity combination being executed at the time of abnormality occurrence, etc.

[0078] By analyzing these abnormality information, the operation and maintenance personnel can more accurately locate the problem and take appropriate measures to repair in time, thereby ensuring the normal operation of the switch cabinet control system.

[0079] Therefore, under any active session category, as long as an exception occurs when the control unit instance is executing the state update activity combination corresponding to the active session category, the exception information needs to be reported to the switch cabinet system operation and maintenance center. These exception information can include the response time of the control unit instance, operation details, etc., which can help operation and maintenance personnel to more accurately understand and handle exceptions.

[0080] Based on the above steps, by obtaining the state update activity combination and its classification index combination of the target switch cabinet control system, and distinguishing between the dominant state activity and the reactive state activity, the activity session category of each state update activity combination is determined, which not only helps operation and maintenance personnel to more effectively manage and monitor the switch cabinet control task, but also timely reports the exception when the control unit instance appears an exception, thereby avoiding or reducing the possible loss. In addition, the exception detection mechanism based on the activity session category can more accurately locate the exception, whether it is a time domain or an operation domain exception can be accurately captured and reported to the switch cabinet system operation and maintenance center. In this way, operation and maintenance personnel can quickly and accurately find the problem and take necessary repair measures to ensure the stable operation of the switch cabinet control system. That is, the present application significantly improves the intelligent degree and operation efficiency of the switch cabinet control system, and also provides a powerful tool for realizing accurate exception detection and rapid fault repair.

[0081] In a possible implementation, step S120 can include:

[0082] Step S121, obtaining the description vector of the X dominant state activities in the X state update activity combinations, generating X dominant state description features, obtaining the description vector of the X reactive state activities in the X state update activity combinations, generating X reactive state description features, obtaining the index category features of the activity classification indexes of the X dominant state activities in the X classification index combinations, generating X first index category description features, and obtaining the index category features of the activity classification indexes of the X reactive state activities in the X classification index combinations, generating X second index category description features.

[0083] Suppose a power switch cabinet control system is running, which has two state update activity combinations (i.e. X = 2). Each state update activity combination includes a dominant state activity and a reactive state activity.

[0084] First, the dominant state activity and reactive state activity in each of the two state update activity combinations can be obtained. For example, the first state update activity combination can involve the dominant state activity "scheduling power according to preset operation rules" and the reactive state activity "adjusting power in real time according to actual power grid load changes"; the second state update activity combination can involve the dominant state activity "periodically checking the state of the switch cabinet" and the reactive state activity "starting the repair program immediately after a fault is found". Then, the dominant state description feature and the reactive state description feature can be generated based on these description vectors. In addition, the index category feature of the activity classification index corresponding to each state activity can also be obtained, and the first index category description feature and the second index category description feature can be generated.

[0085] In step S122, the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features are fused to generate X target feature representations corresponding to the X state update activity combinations.

[0086] This fusion process can involve concatenating all these feature vectors together, or using a more complex method (such as a neural network) to generate a comprehensive feature representation that contains all the relevant information.

[0087] In step S123, based on the X target feature representations, X activity session categories of the X state update activity combinations are determined, each of the X activity session categories reflecting the conversation relationship between the dominant state activity and the reactive state activity in the corresponding one of the X state update activity combinations.

[0088] For example, if the target feature representation of a state update activity combination indicates that its dominant state activity is in progress and the reactive state activity has been completed, it can be assigned to an activity session category called "normal operation". If the dominant state activity fails and needs to be handled immediately, it can be assigned to an activity session category called "abnormal repair". If the system detects an emergency situation, such as the power grid load exceeding a threshold, and the reactive state activity fails to respond in time, it can be assigned to an activity session category called "emergency handling".

[0089] Therefore, through the above steps, the activity session category of each state update activity combination can be determined based on the dominant state description feature, the first index category description feature, the reactive state description feature, and the second index category description feature.

[0090] In a possible implementation, in step S121, obtaining the description vector of the X dominant state activities in the X state update activity combinations and generating the X dominant state description features can include:

[0091] The description vector of the rth dominant state activity in the X dominant state activities is obtained based on the following operations, and the rth dominant state description feature is generated, where r is a positive integer not less than 1 and not greater than X:

[0092] In step S1211, when the rth dominant state activity includes Yr state behaviors, the behavior feature representation of each state behavior in the Yr state behaviors is obtained, and Yr behavior feature representations are generated, where Yr is a positive integer.

[0093] For example, assume that an electric switch cabinet control system is running, which has two state update activity combinations (i.e., X = 2). Each state update activity combination includes a dominant state activity and a reactive state activity. For example, the first dominant state activity can be "scheduling power according to preset operation rules", which includes Y1 = 3 state behaviors: calculating preset operation rules, setting power scheduling parameters, and starting power scheduling. Each state behavior has a behavior feature representation, for example, the behavior feature representation of calculating preset operation rules can include calculation time, algorithm type used, etc. After obtaining the behavior feature representations of the three state behaviors, three behavior feature representations are obtained.

[0094] In step S1212, the Yr behavior feature representations are balanced to generate the rth dominant state description feature. Alternatively, the Yr behavior feature representations are weighted and fused to generate the rth dominant state description feature.

[0095] Next, the three behavior feature representations are balanced or weighted and fused. Balancing can mean that the three feature representations are normalized so that each feature representation has the same influence in the entire description feature; and weighting and fusing can mean that different weights are assigned according to the importance of each state behavior, and then the weighted feature representations are fused together. Regardless of which method is adopted, the first dominant state description feature is finally generated.

[0096] Similarly, the second dominant state activity can be "periodically checking the state of the switch cabinet", which includes Y2 = 2 state behaviors: starting the checking program and analyzing the checking results. After obtaining the behavior feature representations of the two state behaviors, two behavior feature representations are obtained. Then the same balancing or weighting and fusing is performed, and finally the second dominant state description feature is generated.

[0097] Therefore, by the above steps, the corresponding dominant state description feature can be generated based on the behavior feature representation of the state behavior in each dominant state activity.

[0098] In a possible implementation, in step S121, the description vector of the Xth reactive state activity in the X state update activity combination is obtained, and an Xth reactive state description feature is generated, including:

[0099] The description vector of the rth reactive state activity in the X reactive state activities is obtained and the rth reactive state description feature is generated based on the following operations, where r is a positive integer not less than 1 and not greater than X:

[0100] In step S1213, when the rth reactive state activity includes Zr state behaviors, the behavior feature representation of each state behavior in the Zr state behaviors is obtained, and Zr behavior feature representations are generated, where Zr is a positive integer.

[0101] For example, continue to use the power switch cabinet control system as an example to explain these steps. Again, assume that there are two state update activity combinations (i.e., X = 2), each of which includes a dominant state activity and a reactive state activity.

[0102] First, the description vector of each reactive state activity needs to be obtained. For example, the first reactive state activity may be "real-time power adjustment according to actual power grid load changes", which includes Z1 = 3 state behaviors: reading power grid load data, calculating load adjustment value, and executing load adjustment. Each state behavior has a behavior feature representation, such as the behavior feature representation of reading power grid load data, which may include data reading time, amount of data read, etc. After obtaining the behavior feature representations of the 3 state behaviors, 3 behavior feature representations are obtained.

[0103] In step S1214, the Zr behavior feature representations are balanced to generate the rth reactive state description feature. Alternatively, the Zr behavior feature representations are weighted and fused to generate the rth reactive state description feature.

[0104] Next, the 3 behavior feature representations are balanced or weighted and fused. Balancing may mean normalizing the 3 feature representations so that each feature representation has the same influence in the entire description feature; while weight fusion may mean assigning different weights to each state behavior according to its importance, and then fusing the weighted feature representations together. Regardless of which method is taken, the first reactive state description feature is finally generated.

[0105] Similarly, the second reactive state activity is processed: the second reactive state activity can be "start repair procedure immediately after fault is detected", which contains Z2=2 state behaviors: detecting switch cabinet fault, starting repair procedure. After obtaining the behavior feature representation of the two state behaviors, two behavior feature representations are obtained. Then the same balancing processing or weight fusion is performed, and finally the second reactive state description feature is generated.

[0106] Therefore, through the above steps, the corresponding reactive state description feature can be generated based on the behavior feature representation of the state behavior in each reactive state activity.

[0107] In a possible implementation, in step S121, the index category feature of the activity classification index of the X dominant state activities in the X classification index combination is obtained, and X first index category description features are generated, including:

[0108] The index category feature of the activity classification index of the rth dominant state activity in the X dominant state activities is obtained based on the following operations, and the rth first index category description feature is generated, r is a positive integer not less than 1 and not greater than X:

[0109] In step S1215, when the activity classification index of the rth dominant state activity includes an rth index description field, the field feature representation of the rth index description field is obtained, and the rth first index category description feature is the field feature representation of the rth index description field.

[0110] For example, still using the previous power switch cabinet control system as an example. It is assumed that there are two state update activity combinations (i.e., X=2), and each state update activity combination includes a dominant state activity and a reactive state activity.

[0111] First, the index category feature of the activity classification index of each dominant state activity needs to be obtained. For example, the first dominant state activity can be "scheduling power according to preset operation rules", and its activity classification index can include an index description field: "load scheduling". At this time, the field feature representation can include the semantic information of the field, the correlation with other fields, etc. In this example, the first index category description feature of the first dominant state activity is the field feature representation of "load scheduling".

[0112] Similarly, the second dominant state activity is processed: the second dominant state activity can be "periodically check the state of the switch cabinet", and its activity classification index can include an index description field: "state check". Similarly, the feature representation of the field is obtained as the first index category description feature of the second dominant state activity.

[0113] Therefore, by the above steps, the corresponding first index category description feature can be generated based on the index description field of the activity classification index of each dominant state activity.

[0114] In a possible implementation, in step S121, the index category feature of the activity classification index of the X reactive state activities in the X classification index combination is obtained, and X second index category description features are generated, including:

[0115] The index category feature of the activity classification index of the rth reactive state activity in the X reactive state activities is obtained based on the following operations, and the rth second index category description feature is generated, where r is a positive integer not less than 1 and not greater than X:

[0116] In step S1216, when the activity classification index of the rth reactive state activity includes the rth index description field, the field feature representation of the rth index description field is obtained, and the rth second index category description feature is the field feature representation of the rth index description field.

[0117] For example, continue to use the power switch cabinet control system as an example to explain these steps. It is assumed that there are two state update activity combinations (i.e., X = 2), and each state update activity combination includes one dominant state activity and one reactive state activity.

[0118] First, the index category feature of the activity classification index of each reactive state activity needs to be obtained. For example, the first reactive state activity can be "real-time power adjustment according to actual power grid load changes", and its activity classification index can include an index description field "real-time adjustment". At this time, the field feature representation can include the semantic information of the field, the correlation with other fields, etc. In this example, the second index category description feature of the first reactive state activity is the field feature representation of "real-time adjustment".

[0119] Similarly, the second reactive state activity is processed: the second reactive state activity can be "immediately start the repair program after discovering the fault", and its activity classification index can include an index description field "fault repair". Similarly, the feature representation of the field is obtained as the second index category description feature of the second reactive state activity.

[0120] Therefore, by the above steps, the corresponding second index category description feature can be generated based on the index description field of the activity classification index of each reactive state activity.

[0121] In a possible implementation, step S122 can include:

[0122] The rth target feature representation corresponding to the rth state update activity combination in the X state update activity combinations is obtained based on the following operations, the rth state update activity combination includes an rth dominant state activity and an rth reactive state activity, r is a positive integer not less than 1 and not greater than X:

[0123] In step S1221, the rth target feature representation is generated by fusing the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature. The task trigger description vector reflects the trigger node of the target switch cabinet control task. The state update activities in the X state update activity combinations are state update activities parsed from the target switch cabinet control task. The rth dominant state description feature is a description vector of the rth dominant state activity. The rth first index category description feature is an index category feature of the activity classification index of the rth dominant state activity. The rth reactive state description feature is a description vector of the rth reactive state activity. The rth second index category description feature is an index category feature of the activity classification index of the rth reactive state activity.

[0124] When the dimensions of the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are all 1*K, the rth target feature representation with a dimension of 5*K is generated by fusing the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature. K is a positive integer greater than 1.

[0125] For example, these steps are explained using the example of the power switch cabinet control system. It is assumed that there are two state update activity combinations (i.e., X=2), and each state update activity combination includes one dominant state activity and one reactive state activity.

[0126] First, the task trigger description vector reflecting the trigger node of the target switch cabinet control task needs to be obtained. For example, if a target switch cabinet control task is triggered by a timer, the task trigger description vector may include trigger time, trigger condition, and other information.

[0127] Then, the task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are fused to generate an rth target feature representation. In this example, assuming that all these features are 1*K vectors, the fused target feature representation is a 5*K matrix. For example, if K=3, the target feature representation can be represented in the following format:

[0128] [task trigger description vector]

[0129] [dominant state description feature]

[0130] [first index category description feature]

[0131] [reactive state description feature]

[0132] [second index category description feature]

[0133] Each row is a 1*3 vector representing the corresponding feature. In this way, the target feature representation of each state update activity combination is obtained.

[0134] Therefore, through the above steps, the target feature representation of each state update activity combination can be generated based on the task trigger description vector, the dominant state description feature, the first index category description feature, the reactive state description feature, and the second index category description feature.

[0135] In one possible implementation, step S123 can include:

[0136] Based on the following operation, determine the rth activity session category of the rth state update activity combination of the X state update activity combinations based on the rth target feature representation of the X target feature representations, r being a positive integer not less than 1 and not greater than X:

[0137] Step S1231, load the rth target feature representation to an AI neural network to generate M session category classification data corresponding to M activity session categories, the M session category classification data reflecting the possibility that the rth activity session category is each candidate session category of the M set activity session categories, M being a positive integer greater than 1.

[0138] For example, in the example of the power switch cabinet control system, it can be assumed that there are several preset activity session categories (M), such as "normal operation", "abnormal repair", and "emergency treatment". These categories describe the possible execution of a state update activity combination.

[0139] First, the target feature representation is loaded into the AI neural network. For example, the target feature representation of the first status update activity combination is inputted into a pre-trained AI neural network, which is capable of outputting probabilities of M activity conversation categories according to the inputted feature representation. This means that the AI neural network will predict which activity conversation category the status update activity combination is most likely to belong to.

[0140] For example, assume that the AI neural network outputs probabilities of [0.7, 0.2, 0.1], which means that the AI neural network predicts that the first status update activity combination has a 70% probability of being "normal operation", a 20% probability of being "abnormal repair", and a 10% probability of being "emergency handling".

[0141] Step S1232, the rth activity conversation category is determined to be equal to the target activity conversation category among the M set activity conversation categories, in which the rth activity conversation category has the largest probability of being the target activity conversation category among the M conversation category classification data.

[0142] Next, the activity conversation category with the largest probability is selected as the activity conversation category of the status update activity combination. In this example, because "normal operation" has the highest probability (70%), the activity conversation category of the first status update activity combination is determined to be "normal operation".

[0143] The same process is also applied to the second status update activity combination. For example, if the second status update activity combination is predicted to have an 80% probability of being "abnormal repair", then its activity conversation category is determined to be "abnormal repair".

[0144] Therefore, through the above steps, the activity conversation category of each status update activity combination can be determined based on the target feature representation and the prediction of the AI neural network.

[0145] In one possible implementation, step S123 can further include:

[0146] Step S1233, loading the X target feature representations round by round to the AI neural network to generate a corresponding session category classification data sequence for each round target feature representation, the each round target feature representation comprising J target feature representations among the X target feature representations, the corresponding session category classification data sequence for the each round target feature representation comprising J sets of session category classification data, each set of session category classification data among the J sets of session category classification data comprising M session category classification data, the M session category classification data covered by the s-th set of session category classification data reflecting the possibility that the s-th active session category corresponding to the s-th target feature representation is each candidate session category among the M set active session categories, the s-th set of session category classification data corresponding to the s-th target feature representation among the J target feature representations, M being a positive integer greater than 1, J being a positive integer not less than 2 and less than X, and s being a positive integer not less than 1 and not greater than J.

[0147] The s-th active session category among the J active session categories is determined based on the s-th set of session category classification data among the session category classification data sequence by using the following operation:

[0148] Step S1234, determining the s-th active session category to be equal to a target active session category among the M set active session categories, the s-th active session category being the target active session category with the largest possibility among the M session category classification data covered by the s-th set of session category classification data.

[0149] In the example of the power switch cabinet control system, assume that there are three state update activity combinations (i.e., X = 3), each including a dominant state activity and a reactive state activity. In addition, three preset active session categories (M) are set, such as "normal operation", "abnormal repair" and "emergency treatment". Now, the steps are explained as follows:

[0150] Assume that only two target feature representations of state update activity combinations are processed in each round (i.e., J = 2). In the first round, the target feature representations of the first two state update activity combinations are input to the AI neural network to obtain two sets of session category classification data, each containing three elements (corresponding to the three preset active session categories). For example, the first set of data may be [0.6, 0.3, 0.1], indicating that the probability of the first state update activity combination being "normal operation" is 60%, "abnormal repair" is 30%, and "emergency treatment" is 10%.

[0151] In the second round, the target feature representation of the third state update activity combination is input to the AI neural network to obtain another set of session category classification data. In this way, a session category classification data sequence is obtained.

[0152] Next, the corresponding active session category is determined for each set of session category classification data. The active session category with the highest probability in each set of data is selected as the active session category for the state update activity combination. In the first round, because the probability of "normal operation" is the highest (60%), the active session category for the first state update activity combination is determined as "normal operation".

[0153] Similarly, in the second round, the active session category with the highest probability is also selected as the active session category for the third state update activity combination.

[0154] In this way, the active session category of each state update activity combination is determined round by round based on the target feature representation and the prediction of the AI neural network.

[0155] It is worth noting that the AI neural network can be trained through supervised learning. First, a training sample set containing a large amount of historical data is needed, and each training sample contains a feature representation of a state update activity combination and a corresponding active session category label.

[0156] Here are some specific steps in the training process:

[0157] First, the original data needs to be preprocessed, which may include steps such as data cleaning, handling missing values, encoding categorical variables, etc. For example, it may be necessary to convert the descriptions of the dominant state activity and reactive state activity into numerical features so that the neural network can process them.

[0158] Next, features that are useful for predicting active session categories need to be selected or extracted. These features may include the type of dominant state activity, the execution state, the response time of reactive state activity, etc.

[0159] Then, the structure of the neural network needs to be designed, including the number of hidden layers, the number of nodes in each layer, the activation function, etc. In addition, a suitable loss function and optimizer need to be selected.

[0160] During the training phase, the training samples in the training sample set are input into the neural network, which will predict the active session category based on the target feature representation of each training sample, and then calculate the loss based on the predicted result and the real active session category label. Then, the optimizer is used to adjust the network parameters to minimize the loss.

[0161] And the performance of the model needs to be evaluated on a separate validation set, and necessary parameter adjustments need to be made. Finally, the final performance of the neural network is verified on the test set.

[0162] After the training is completed, an AI neural network that can predict the activity session category of a state update activity combination is obtained. When a new state update activity combination appears, its target feature representation can be input into the neural network to obtain the corresponding activity session category.

[0163] For example, in a possible implementation, step S110 can include:

[0164] Step S111, parsing H state update activities and H activity classification indexes that exist in a unique mapping relationship in the target switch cabinet control task, the H activity classification indexes including an activity classification index of each of the H state update activities, each of the H state update activities including an element in the target switch cabinet control task.

[0165] Step S112, forming X state update activity combinations from two state update activities in an X group of the H state update activities, each of the X state update activity combinations including two state update activities in a group of the H state update activities, the state update activity with an active trigger state in the two state update activities being the dominant state activity, and the state update activity with a passive trigger state being the reactive state activity.

[0166] Step S113, determining, in the H activity classification indexes, the activity classification indexes of the dominant state activity and the reactive state activity covered in each of the X state update activity combinations, to generate X classification index combinations.

[0167] For example, in a possible implementation, step S110 can include:

[0168] For example, it can be assumed that a target switch cabinet control task contains H state update activities and corresponding H activity classification indexes. For example, these state update activities can include load scheduling, power grid monitoring, fault repair, etc., and each activity has an activity classification index corresponding thereto.

[0169] Firstly, H state update activities and corresponding H activity classification indexes are parsed from the target switch cabinet control task. For example, if an element in the target switch cabinet control task is "performing load scheduling", the corresponding state update activity is "load scheduling", and its activity classification index can be a code or a label indicating whether the activity is actively triggered or passively triggered, and its completion status, etc.

[0170] Then, the H state update activities are divided into X groups (each group contains two state update activities), and the groups are constructed into X state update activity combinations. For example, the first state update activity combination can include "load scheduling" (a leading state activity) and "grid monitoring" (a reactive state activity). The two activities are combined to form a state update activity combination.

[0171] Next, the activity classification indexes of the leading state activity and the reactive state activity covered by each state update activity combination are determined from the H activity classification indexes, and X classification index combinations are generated. For example, for the first state update activity combination mentioned above, the activity classification indexes corresponding to "load scheduling" and "grid monitoring" are selected to form a classification index combination.

[0172] In this process, the conversation relationship between the leading state activity and the reactive state activity represented by each activity session category is one of the preset M conversation relationships, such as "normal operation", "abnormal repair" and "emergency treatment". Of course, there can also be no conversation relationship, such as two activities being completely independent and having no interaction.

[0173] Further, Figure 2 A hardware structure schematic diagram of a state monitoring system 100 for implementing the method provided by the embodiments of the application is shown. As shown in the figure, Figure 2 The state monitoring system 100 includes a memory 111, a storage controller 112, a processor 113, a peripheral interface 114, an input / output unit 115, an audio unit 116, a display unit 117, and a radio frequency unit 118.

[0174] The memory 111, the storage controller 112, the processor 113, the peripheral interface 114, the input / output unit 115, the audio unit 116, the display unit 117, and the radio frequency unit 118 are directly or indirectly electrically connected to each other to realize the transmission or interaction of data. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines.

[0175] The memory 111 can be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electric Erasable Programmable Read-Only Memory (EEPROM), etc. The memory 111 is configured to store a program, and the processor 113 is configured to execute the program after receiving an execution instruction. The processor 113 and other possible components access the memory 111 under control of the memory controller 112.

[0176] The processor 113 can be an integrated circuit chip having a processing capability. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc. The processor can also be a Digital Signal Processor (DSP), an Application-Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The processor can implement or execute the disclosed methods, steps, and logical block diagrams in the embodiments of the present application. The general-purpose processor can be a microprocessor, or the processor can also be any conventional processor.

[0177] The peripheral interface 114 is configured to couple various input / output devices to the processor 113 and the memory 111. In some embodiments, the peripheral interface 114, the processor 113, and the memory controller 112 can be implemented in a single chip. In other embodiments, they can be implemented by independent chips, respectively.

[0178] The input / output unit 115 is configured to provide input data for a user to interact with the state monitoring system 100. The input / output unit 115 can be, but is not limited to, a mouse and a keyboard, etc.

[0179] The audio unit 116 provides an audio interface for a user, which can include one or more microphones, one or more speakers, and audio circuitry.

[0180] The display unit 117 provides an interactive interface (e.g., a user operation interface) or displays image data between the state monitoring system 100 and a user. In the embodiment, the display unit 117 can be a liquid crystal display or a touch display. If it is a touch display, it can be a capacitive touch screen or a resistive touch screen supporting single-point and multi-point touch operations. Supporting single-point and multi-point touch operations means that the touch display can sense touch operations generated from one or more positions on the touch display and transfer the sensed touch operations to a processor for calculation and processing.

[0181] The radio frequency unit 118 is configured to receive and send radio wave signals (e.g., electromagnetic waves) to realize the conversion between radio waves and electrical signals, so as to realize the communication between the state monitoring system 100 and the network 300 or other communication devices.

[0182] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above-mentioned embodiments of the present application are described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims 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.

[0183] Each of the embodiments in the embodiments of the present application is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the above different embodiments, since they are basically similar to the method embodiments, they are described more simply, and the relevant parts can refer to the part of the method embodiments.

[0184] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or by program instructions to instruct related hardware to complete. The above-mentioned program can be stored in a computer readable storage medium, and the above-mentioned storage medium can be a read-only memory, a magnetic disk or an optical disk.

Claims

1. A switch cabinet status monitoring method based on artificial intelligence, characterized in that: The method comprises: Obtaining X state update activity combinations and X classification index combinations of a target switchgear control system, where X is a positive integer greater than 1, each of the X state update activity combinations including a dominant state activity and a reactive state activity of a pending activity session category, the X classification index combinations including activity classification indexes of each dominant state activity and each reactive state activity in the X state update activity combinations, each state update activity combination including a dominant state activity in an active triggering state and a reactive state activity in a passive triggering state, the dominant state activity including a dominant element in a target switchgear control task, the dominant element including a preset operation rule or an instruction sequence, the reactive state activity including a reactive element in the target switchgear control task, the reactive element including a feedback mechanism or an adaptive strategy; Determining X activity session categories of the X state update activity combinations based on X dominant state description features, X first index category description features, X reactive state description features, and X second index category description features, wherein the X dominant state description features are description vectors of the X dominant state activities in the X state update activity combinations, the X first index category description features are index category features of the X dominant state activities, the X reactive state description features are description vectors of the X reactive state activities in the X state update activity combinations, the X first index category description features are index category features of the X dominant state activities, and the X second index category description features are index category features of the X reactive state activities; When an abnormality occurs in a control unit instance in a target node, as represented by any one of the X active session categories, target instance abnormality information is reported to the switchgear system operation and maintenance center. The abnormality occurs in the control unit instance in the target node, including an abnormality in the time domain and / or operation domain of the control unit instance response. The target instance abnormality information reflects that the abnormality occurs in the control unit instance in the target node. The step of obtaining X state update activity combinations and X classification index combinations of the target switchgear control system includes: Parsing, in the target switchgear control task, H state update activities and H activity classification indexes that have a unique mapping relationship, wherein the H activity classification indexes include an activity classification index of each of the H state update activities, and each of the H state update activities includes an element of the target switchgear control task; Combining two status update activities from X groups of the H status update activities into the X status update activity combinations, where each of the X status update activity combinations includes two status update activities from one group of the H status update activities, wherein the status update activity in the one group that has an active triggering state is the dominant status activity, and the status update activity in the one that has a passive triggering state is the reactive status activity; Determining, from the H activity classification indexes, the activity classification indexes of the dominant state activity and the reactive state activity covered in each state update activity combination in the X state update activity combinations, to generate the X classification index combinations; Among them, the session relationship between the dominant state activity and the reactive state activity represented by each active session category in the X active session categories is one of the set M session relationships, M is a positive integer greater than 1, and the M session relationships include M-1 set session relationships and no session relationship.

2. The switch cabinet status monitoring method based on artificial intelligence according to claim 1 is characterized in that: The determining, based on the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features, the X active session categories of the X state update activity combinations includes: Obtaining description vectors of X dominant state activities in the X state update activity combinations to generate X dominant state description features; obtaining description vectors of X reactive state activities in the X state update activity combinations to generate X reactive state description features; obtaining index category features of activity classification indexes of the X dominant state activities in the X classification index combinations to generate X first index category description features; and obtaining index category features of activity classification indexes of the X reactive state activities in the X classification index combinations to generate X second index category description features; Fusing the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features to generate X target feature representations corresponding to the X state update activity combinations; Based on the X target feature representations, X activity session categories of the X state update activity combinations are determined, each of the X activity session categories reflecting a session relationship between the dominant state activity and the reactive state activity in a corresponding one of the X state update activity combinations.

3. The switch cabinet status monitoring method based on artificial intelligence according to claim 2 is characterized in that: The obtaining of description vectors of X dominant state activities in the X state update activity combinations to generate X dominant state description features includes: The description vector of the rth dominant state activity among the X dominant state activities is obtained based on the following operation to generate the rth dominant state description feature, where r is a positive integer not less than 1 and not greater than X: When the r-th dominant state activity includes Yr state behaviors, obtaining a behavior feature representation of each of the Yr state behaviors to generate Yr behavior feature representations, where Yr is a positive integer; Performing equalization processing on the Yr behavioral feature representations to generate the rth dominant state description feature; or performing weight fusion on the Yr behavioral feature representations to generate the rth dominant state description feature.

4. The switch cabinet status monitoring method based on artificial intelligence according to claim 2 is characterized in that: The obtaining of description vectors of X reactive state activities in the X state update activity combinations to generate X reactive state description features includes: Obtain a description vector of the rth reactive state activity among the X reactive state activities based on the following operation, and generate an rth reactive state description feature, where r is a positive integer not less than 1 and not greater than X: When the rth reactive state activity includes Zr state behaviors, obtaining a behavior feature representation of each of the Zr state behaviors to generate Zr behavior feature representations, where Zr is a positive integer; The Zr behavioral feature representations are balanced to generate the rth reactive state description feature; or, weighted fusion is performed on the Zr behavioral feature representations to generate the rth reactive state description feature.

5. The switch cabinet status monitoring method based on artificial intelligence according to claim 2 is characterized in that: The step of obtaining index category features of the activity classification indexes of the X dominant state activities in the X classification index combinations to generate X first index category description features includes: The index category feature of the activity classification index of the rth dominant state activity among the X dominant state activities is obtained based on the following operation to generate the rth first index category description feature, where r is a positive integer not less than 1 and not greater than X: When the activity classification index of the rth dominant state activity includes the rth index description field, the field feature representation of the rth index description field is obtained, and the rth first index category description feature is the field feature representation of the rth index description field.

6. The switch cabinet status monitoring method based on artificial intelligence according to claim 2 is characterized in that: The acquiring index category features of the activity classification indexes of the X reactive state activities in the X classification index combinations to generate X second index category description features includes: The index category feature of the activity classification index of the rth reactive state activity among the X reactive state activities is obtained based on the following operation, and the rth second index category description feature is generated, where r is a positive integer not less than 1 and not greater than X: When the activity classification index of the rth reactive state activity includes the rth index description field, the field feature representation of the rth index description field is obtained, and the rth second index category description feature is the field feature representation of the rth index description field.

7. The switch cabinet status monitoring method based on artificial intelligence according to claim 2 is characterized in that: The fusing the X dominant state description features, the X first index category description features, the X reactive state description features, and the X second index category description features to generate X target feature representations corresponding to the X state update activity combinations includes: The rth target feature representation corresponding to the rth state update activity combination among the X state update activity combinations is obtained based on the following operation, where the rth state update activity combination includes the rth dominant state activity and the rth reactive state activity, and r is a positive integer not less than 1 and not greater than X: The task trigger description vector, the rth dominant state description feature, the rth first index category description feature, the rth reactive state description feature, and the rth second index category description feature are integrated to generate the rth target feature representation, wherein the task trigger description vector reflects the trigger node of the target switchgear control task, the state update activities in the X state update activity combinations are state update activities parsed from the target switchgear control task, the rth dominant state description feature is the description vector of the rth dominant state activity, the rth first index category description feature is the index category feature of the activity classification index of the rth dominant state activity, the rth reactive state description feature is the description vector of the rth reactive state activity, and the rth second index category description feature is the index category feature of the activity classification index of the rth reactive state activity; Among them, when the dimension of the task trigger description vector, the dimension of the r-th dominant state description feature, the dimension of the r-th first index category description feature, the dimension of the r-th reactive state description feature and the dimension of the r-th second index category description feature are all 1*K, the task trigger description vector, the r-th dominant state description feature, the r-th first index category description feature, the r-th reactive state description feature and the r-th second index category description feature are fused to generate the r-th target feature representation with a dimension of 5*K, where K is a positive integer greater than 1.

8. The switch cabinet status monitoring method based on artificial intelligence according to claim 2, characterized in that: The determining, based on the X target feature representations, X activity session categories of the X status update activity combinations includes: Determine an rth active session category of an rth status update activity combination among the X status update activity combinations based on an rth target feature representation among the X target feature representations based on the following operations, where r is a positive integer not less than 1 and not greater than X: Loading the rth target feature representation into an AI neural network to generate M session category classification data corresponding to the M active session categories, wherein the M session category classification data reflect the likelihood that the rth active session category is each candidate session category among the M set active session categories, where M is a positive integer greater than 1; The rth active session category is determined to be equal to a target active session category among the M set active session categories, and among the M session category classification data, the rth active session category has the highest probability of being the target active session category.

9. The switch cabinet status monitoring method based on artificial intelligence according to claim 2, characterized in that: The determining, based on the X target feature representations, X activity session categories of the X status update activity combinations includes: The X target feature representations are loaded into the AI ​​neural network in rounds to generate a conversation category classification data sequence corresponding to the target feature representations of each round, wherein the target feature representations of each round include J target feature representations from the X target feature representations, and the conversation category classification data sequence corresponding to the target feature representations of each round includes J groups of conversation category classification data, each group of conversation category classification data in the J groups of conversation category classification data includes M conversation category classification data, and the M conversation category classification data covered by the s-th group of conversation category classification data reflect the likelihood that the s-th active conversation category corresponding to the s-th target feature representation is each candidate conversation category from the M set active conversation categories, and the s-th group of conversation category classification data corresponds to the s-th target feature representation in the J target feature representations, where M is a positive integer greater than 1, J is a positive integer not less than 2 and less than X, and s is a positive integer not less than 1 and not greater than J. Determine the sth active session category among J active session categories based on the sth group of session category classification data in the sequence of session category classification data by using the following operations: The sth active session category is determined to be equal to the target active session category among the M set active session categories. Among the M session category classification data covered by the sth group of session category classification data, the sth active session category is most likely to be the target active session category.

10. A status monitoring system, characterized in that: The state monitoring system includes a processor and a readable storage medium, wherein the readable storage medium stores a program, and when the program is executed by the processor, the switch cabinet state monitoring method based on artificial intelligence according to any one of claims 1 to 9 is implemented.

Images