Energy industry internet container arrangement safety management system and arrangement method

Abstract

The invention discloses an energy industry internet container arrangement safety management system and an arrangement method. The system comprises an environment sequence construction module which is used for constructing an environment sequence; the service sequence construction module is used for constructing a service sequence; the environment influence prediction module is used for obtaining an environment influence score through an environment prediction model; the business influence prediction module is used for obtaining a business influence score through a business prediction model; the container early warning arrangement module is used for performing weighted summation according to the environmental influence score and the business influence score to obtain a comprehensive influence score, and generating a container early warning arrangement strategy according to the comprehensive influence score; according to the method, the environment prediction model and the service prediction model are used for performing time sequence analysis on the environment data and the service data respectively, so that the functions of load change prediction and container automatic arrangement are realized, and the service response delay is reduced.

Description

Technical Field

[0001] The present invention relates to the field of load forecasting technology, and in particular to an energy industrial Internet container orchestration security management system and an orchestration method. Background Art

[0002] With the rapid development of the energy industrial internet, more and more energy companies are implementing modularization and high availability of their business systems through microservices architecture and containerized deployment. Container orchestration enables automated deployment, scaling, load balancing, and fault-tolerance management of distributed applications, thereby improving system flexibility and resource utilization.

[0003] However, in the actual application scenarios of the energy industrial Internet, the load pressure faced by the system is uncertain and sudden. For example, the power grid dispatching system may face large-scale traffic requests in severe weather. The current mainstream container orchestration mechanism mainly relies on fixed threshold rule settings (such as CPU and memory usage) to trigger scaling adjustments, or relies on manual intervention. This type of "passive response" mode often has the risk of response delays, scheduling lags, and even service interruptions when facing large-scale concurrent requests or traffic peaks. Therefore, the existing container orchestration mechanism lacks an early warning mechanism, making it difficult to predict load changes in advance based on business operation trends and changes in the external environment, resulting in service response delays. Summary of the Invention

[0004] In view of this, the embodiment of the present invention provides an energy industry Internet container orchestration security management system and orchestration method

[0005] A first aspect of the present invention provides an energy industrial internet container orchestration safety management system, comprising an environment sequence construction module, a business sequence construction module, an environment impact prediction module, a business impact prediction module, and a container early warning orchestration module;

[0006] The environment sequence construction module is used to collect environment data at a first preset time interval a1 within a first preset time period A1, and perform normalization processing to form multiple sequence units to construct an environment sequence;

[0007] The service sequence construction module is used to collect service data according to the second preset time interval b1 within the second preset time period B1, and perform normalization processing to form multiple sequence units to construct a service sequence;

[0008] The environmental impact prediction module is used to input the environmental sequence into the trained environmental prediction model and output the environmental impact score in the future time period Q;

[0009] The business impact prediction module is used to input the business sequence into the trained business prediction model and output the business impact score in the future time period Q;

[0010] The container early warning orchestration module is used to perform a weighted summation based on the environmental impact score and the business impact score in the future time period Q to obtain a comprehensive impact score, and generate a container early warning orchestration strategy based on the comprehensive impact score.

[0011] Furthermore, the number of sequence units in the environment sequence is the ratio of the first preset time period A1 to the first preset time interval a1; the number of sequence units in the service sequence is the ratio of the second preset time period B1 to the second preset time interval b1.

[0012] Furthermore, the environment prediction model is composed of M first hidden layers and a first classifier, wherein the number M of the first hidden layers is the same as the number of sequence units of the environment sequence;

[0013] The mth sequence unit of the mth first hidden layer inputs the environment sequence and outputs the first update vector, 1≤m≤M;

[0014] The first update vectors output by the M first hidden layers are input into the first classifier, and the environmental impact score in the future time period Q is represented by the classification space of the first classifier.

[0015] Furthermore, the calculation formula of the first hidden layer includes:

[0016] F m，1 =PReLU(h m ×W m，1 +b m，1 );

[0017] h m =Swish(X m ×W m，2 +h m-1 ×W m，3 +b m，2 );

[0018] Among them F m，1 represents the first update vector of the output of the mth first hidden layer, X m represents the mth sequence unit of the environment sequence input to the mth first hidden layer, h m and h m-1 They represent the hidden state vectors of the mth and m-1th first hidden layers respectively, and the dimension values ​​of h0 are all assigned to 0, W m，1 、W m，2 and W m，3They represent the first weight parameter, the second weight parameter and the third weight parameter of the mth first hidden layer, respectively, and b m，1 and b m，2 They represent the first bias parameter and the second bias parameter of the mth first hidden layer, PReLU represents the PReLU activation function, and Swish represents the Swish activation function.

[0019] Furthermore, the business prediction model is composed of N second hidden layers and second classifiers, wherein the number N of the second hidden layers is the same as the number of sequence units of the business sequence;

[0020] The nth second hidden layer inputs the nth sequence unit of the business sequence and outputs the second update vector, 1≤n≤N;

[0021] The second update vectors output by the N second hidden layers are input into the second classifier, and the business impact score in the future time period Q is represented by the classification space of the second classifier.

[0022] Furthermore, the calculation formula of the second hidden layer includes:

[0023] The reset gate r of the nth second hidden layer n The calculation formula is as follows:

[0024]

[0025] Where V n represents the nth sequence unit of the business sequence input to the nth second hidden layer, F n-1，2 represents the second update vector of the n-1th second hidden layer output, F 0，2 The dimension values ​​of are all assigned to 0. and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the reset gate of the nth second hidden layer, respectively. Sigmoid represents the Sigmoid activation function.

[0026] The update gate z of the nth second hidden layer n The calculation formula is as follows:

[0027]

[0028] in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the update gate of the nth second hidden layer respectively;

[0029] Candidate hidden states for the nth second hidden layer The calculation formula is as follows:

[0030]

[0031] in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the candidate hidden state of the nth second hidden layer, respectively, and tanh represents the hyperbolic tangent function;

[0032] The second update vector F output by the nth second hidden layer n，2 The calculation formula is as follows:

[0033]

[0034] Where ⊙ represents point-by-point multiplication.

[0035] Furthermore, the environment prediction model and the business prediction model are both trained using preset training samples, and the sample labels of the training samples are obtained through manual evaluation and annotation;

[0036] The environmental impact score output by the environmental prediction model ranges from 0 to 10;

[0037] The business impact score output by the business prediction model ranges from 0 to 10.

[0038] Furthermore, the weight coefficients corresponding to the environmental impact score and the business impact score obtained by weighted summation to obtain the comprehensive impact score are preset parameters;

[0039] When the comprehensive score is less than the first score, there is no need to adjust the container;

[0040] When the comprehensive score is between a first score and a second score, performing hot standby processing on the container in advance; the first score is less than the second score;

[0041] When the comprehensive score is greater than the second score, the container is immediately expanded.

[0042] A second aspect of the present invention discloses a method for orchestrating containers in the energy industry internet, comprising the following steps:

[0043] In a first preset time period A1, environmental data is collected according to a first preset time interval a1, and normalized to form a plurality of sequence units to construct an environmental sequence;

[0044] In a second preset time period B1, service data is collected according to the second preset time interval b1, and normalized to form a plurality of sequence units to construct a service sequence;

[0045] Input the environmental sequence into the trained environmental prediction model and output the environmental impact score in the future time period Q;

[0046] Input the business sequence into the trained business prediction model and output the business impact score in the future time period Q;

[0047] A weighted sum of the environmental impact score and the business impact score in the future time period Q is performed to obtain a comprehensive impact score, and a container early warning orchestration strategy is generated based on the comprehensive impact score.

[0048] Furthermore, the number of sequence units in the environment sequence is the ratio of the first preset time period A1 to the first preset time interval a1; the number of sequence units in the service sequence is the ratio of the second preset time period B1 to the second preset time interval b1.

[0049] The present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the above method.

[0050] The embodiments of the present invention have the following beneficial effects: The present invention provides an energy industrial Internet container orchestration security management system and orchestration method, which respectively performs time series analysis on environmental data and business data through an environmental prediction model and a business prediction model, thereby realizing the functions of load change prediction and automatic container orchestration, thereby reducing service response delays.

[0051] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0053] Figure 1 This is a schematic diagram of an energy industry internet container orchestration security management system of the present invention;

[0054] Figure 2 is a flow chart for generating an environmental impact score according to the present invention;

[0055] Figure 3 This is a flow chart for generating a business impact score according to the present invention.

[0056] Reference numerals: environment sequence construction module 101 , business sequence construction module 102 , environment impact prediction module 103 , business impact prediction module 104 , container warning orchestration module 105 . DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0058] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar terms used in one or more embodiments of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "include" or "comprise" mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0059] like Figure 1 As shown, the first embodiment of the present invention discloses an energy industrial Internet container orchestration security management system, including an environment sequence construction module, a business sequence construction module, an environment impact prediction module, a business impact prediction module and a container early warning orchestration module;

[0060] The environment sequence construction module is used to collect environment data at first preset time intervals a1 within a first preset time period A1, and perform normalization processing to form multiple sequence units to construct an environment sequence;

[0061] The service sequence construction module is used to collect service data at a second preset time interval b1 within a second preset time period B1, and perform normalization processing to form a plurality of sequence units to construct a service sequence;

[0062] The environmental impact prediction module is used to input the environmental sequence into the trained environmental prediction model and output the environmental impact score in the future time period Q;

[0063] The business impact prediction module is used to input the business sequence into the trained business prediction model and output the business impact score in the future time period Q;

[0064] The container early warning orchestration module is used to perform a weighted summation based on the environmental impact score and the business impact score in the future time period Q to obtain a comprehensive impact score, and generate a container early warning orchestration strategy based on the comprehensive impact score.

[0065] In this embodiment of the present invention, the duration of the future time period Q is a custom parameter. For example, the duration of the future time period Q is set to 1 hour. Short-term prediction ensures high system availability. Furthermore, the container warning orchestration policy generated by the container warning orchestration module in this embodiment of the present invention has a priority relationship with the manually set container orchestration rules. The manually set container orchestration rules have a higher priority than the container warning orchestration policy generated by the container warning orchestration module to prevent conflicts between container orchestration policies.

[0066] The following describes the generation process of the environmental impact score and the business impact score according to the embodiment of the present invention respectively.

[0067] Environmental impact rating: Figure 2 As shown, in the embodiment of the present invention, the environmental impact score first collects environmental data through the environmental sequence construction module to construct an environmental sequence; then uses the trained environmental prediction model to make predictions based on the environmental sequence, and the prediction result is the environmental impact score output by the model.

[0068] In one embodiment of the present invention, the environmental data includes: temperature, rainfall, wind speed, air pressure, schedule (such as whether it is a holiday), month, and power load.

[0069] In one embodiment of the present invention, the length of the environmental sequence is equal to the ratio of the first preset time period A1 to the first preset time interval a1, wherein the first preset time period A1 and the first preset time interval a1 are both custom parameters. For example, when the first preset time period A1 is set to 7 days and the first preset time interval a1 is set to 1 day, the length of the environmental sequence is 7, comprising 7 sequence units, each of which records a set of collected environmental data.

[0070] In terms of the environment prediction model, the environment prediction model used in the embodiment of the present invention is composed of M first hidden layers and 1 first classifier, wherein the number M of the first hidden layers is the same as the length of the environment sequence;

[0071] The mth first hidden layer inputs the mth sequence unit of the environment sequence and outputs the first update vector, where the number of dimensions of the first update vector is a custom parameter, 1≤m≤M;

[0072] The first update vector output by the Mth first hidden layer is input into the first classifier, and the classification space of the first classifier represents the environmental impact score in the future time period Q.

[0073] In one embodiment of the present invention, the calculation formula of the mth first hidden layer includes:

[0074] F m，1 =PReLU(h m ×W m，1 +b m，1 );

[0075] h m =Swish(X m ×W m，2 +h m-1 ×W m，3 +b m，2 );

[0076] Among them F m，1 represents the first update vector of the output of the mth first hidden layer, X m represents the mth sequence unit of the environment sequence input to the mth first hidden layer, h m and h m-1 They represent the hidden state vectors of the mth and m-1th first hidden layers respectively. The number of dimensions of the hidden state vector is a custom parameter. The dimension values ​​of h0 are all assigned to 0. W m，1 、W m，2 and W m，3 They represent the first weight parameter, the second weight parameter and the third weight parameter of the mth first hidden layer, respectively, and b m，1 and b m，2 They represent the first bias parameter and the second bias parameter of the mth first hidden layer, PReLU represents the PReLU activation function, and Swish represents the Swish activation function.

[0077] It should be noted that, assuming that the number of dimensions of the first update vector is set to 32 and the number of dimensions of the hidden state vector is set to 16, the first weight parameter needs to be designed as a matrix of 16×32 size, and the first bias parameter needs to be designed as a vector of 1×32 size. Assuming that the number of dimensions of a sequence unit of the environment sequence is 7, the second weight parameter needs to be designed as a matrix of 7×16 size, the third weight parameter needs to be designed as a matrix of 16×16 size, and the second bias parameter needs to be designed as a vector of 1×16 size.

[0078] In one embodiment of the present invention, sample labels of training samples for training an environmental prediction model are obtained through manual evaluation and annotation, where the value range of the environmental impact score is between 0 and 10, and the environmental data of the next time period can be used as the sample labels of pre-training samples for pre-training.

[0079] Business Impact Rating: Figure 3As shown, in the embodiment of the present invention, the business impact score first collects environmental data through the business sequence construction module to construct a business sequence; then uses the trained business prediction model to make predictions based on the business sequence, and the prediction result is the business impact score output by the model.

[0080] In one embodiment of the present invention, the business data includes: the number of requests per unit time, CPU load, memory usage, network bandwidth, average response time, maximum response time, etc.

[0081] In one embodiment of the present invention, the length of a service sequence is equal to the ratio of the second preset time period B1 to the second preset time interval b1, where both the second preset time period B1 and the second preset time interval b1 are custom parameters. For example, when the second preset time period B1 is set to 30 days and the second preset time interval b1 is set to 1 day, the length of the service sequence is 30, comprising 30 sequence units, each of which records a set of collected service data.

[0082] In terms of the business prediction model, the business prediction model used in the embodiment of the present invention is composed of N second hidden layers and 1 second classifier, wherein the number N of the second hidden layers is the same as the length of the business sequence;

[0083] The nth second hidden layer inputs the nth sequence unit of the business sequence and outputs a second update vector, where the number of dimensions of the second update vector is a custom parameter, 1≤n≤N;

[0084] The second update vector output by the Nth second hidden layer is input into the second classifier, and the classification space of the second classifier represents the business impact score in the future time period Q.

[0085] In one embodiment of the present invention, the calculation formula of the nth second hidden layer includes:

[0086] The reset gate r of the nth second hidden layer n The calculation formula is as follows:

[0087]

[0088] Where V n represents the nth sequence unit of the business sequence input to the nth second hidden layer, F n-1，2 represents the second update vector of the n-1th second hidden layer output, F 0，2 The dimension values ​​of are all assigned to 0. and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the reset gate of the nth second hidden layer, respectively. Sigmoid represents the Sigmoid activation function.

[0089] The update gate z of the nth second hidden layer n The calculation formula is as follows:

[0090]

[0091] in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the update gate of the nth second hidden layer respectively;

[0092] Candidate hidden states for the nth second hidden layer The calculation formula is as follows:

[0093]

[0094] in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the candidate hidden state of the nth second hidden layer, respectively, and tanh represents the hyperbolic tangent function;

[0095] The second update vector F output by the nth second hidden layer n，2 The calculation formula is as follows:

[0096]

[0097] Where ⊙ represents point-by-point multiplication.

[0098] In one embodiment of the present invention, sample labels of training samples for training a business prediction model are obtained through manual evaluation and annotation, where the value range of the business impact score is between 0 and 10, and the business data of the next time period can be used as the sample labels of pre-training samples for pre-training.

[0099] It should be noted that the time period for collecting environmental data and business data in the embodiment of the present invention can be scattered in any period of 24 hours. For example, collecting at the same time every day (such as 8 am, 8 pm, etc.) will help compare and analyze data across days, and regular collection can maintain the continuity and stability of the data, and improve the prediction accuracy of the model. It will not be elaborated here.

[0100] It should be noted that the normalization method adopted by the environment sequence construction module and the business sequence construction module in the embodiment of the present invention can be the Min-Max method or the Z-Score method, etc., to eliminate the dimensional differences between the data, thereby improving the generalization ability and robustness of the model.

[0101] It should be noted that the weight parameters and bias parameters in the environmental prediction model and the business prediction model are all learnable hyperparameters, and the reverse update of the parameters can be achieved through the gradient descent algorithm, which will not be elaborated here; in addition, the first classifier and the second classifier are both constructed based on the multi-layer perceptron, and the activation functions of both can be set to the ReLU activation function or the PReLU activation function, so as to achieve regression prediction of the environmental impact score and the business impact score, and the addition of pre-training can reduce the number of training samples in actual training to a certain extent and accelerate the convergence of the model.

[0102] After obtaining the environmental impact score and the business impact score, the embodiment of the present invention performs weighted summation to obtain a comprehensive impact score through the container early warning orchestration module, where the weight coefficients corresponding to the environmental impact score and the business impact score are both custom parameters.

[0103] Among them, when the comprehensive score is less than the first score, there is no need to adjust the container; when the comprehensive score is between the first score and the second score, the container is hot-standbyed in advance; when the first score is less than the second score; when the comprehensive score is greater than the second score, the container is immediately expanded.

[0104] For example, the first score can be set to 7, the second score to 8.5, the weight coefficient corresponding to the environmental impact score to 0.3, and the weight coefficient corresponding to the business impact score to 0.7. In this embodiment, if the comprehensive score is less than 7, no container adjustment is required; if the comprehensive score is greater than or equal to 7 and less than 8.5, containers are preemptively hot-backed up; and if the comprehensive score is greater than or equal to 8.5 and less than 10, capacity expansion is performed immediately.

[0105] It should be noted that hot standby containers refer to starting some container instances in advance in the container orchestration system and keeping them in a "standby" state. The purpose of this is to prepare for the upcoming traffic peak in advance when the system predicts a possible load increase, but it does not process actual user requests until the load is overloaded. Immediate capacity expansion will not only start the preheated container, but also directly expand resources, adding more container instances to carry the current traffic load, such as adding additional memory, etc., and the expanded container instances will be dynamically included in the load balancer until the load falls back to a safe range. I will not go into details here.

[0106] It should be noted that the intervals and thresholds are set for ease of comparison. The threshold size depends on the amount of sample data and the cardinality set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless numerical calculations. These formulas are derived from software simulations of the most recent real-world conditions using large amounts of data. The preset parameters in these formulas are set by those skilled in the art based on actual conditions.

[0107] The following are application examples of the embodiments of the present invention in various industrial scenarios:

[0108] In the UHV power transmission and transformation monitoring scenario, this embodiment of the present invention sets A1 = 5 minutes and a1 = 10 seconds to collect environmental data such as temperature, humidity, and electromagnetic radiation from substation equipment; B1 = 3 minutes and b1 = 30 seconds to collect business metrics such as the power dispatch instruction issuance rate and packet retransmission rate. The environmental prediction model uses a 4-hidden layer GRU architecture to process temperature and humidity sequences, outputting an environmental impact score of 9.2. The business model processes network latency data using an LSTM with a reset gate structure, achieving a business score of 8.5. A weighted calculation of 0.6:0.4 yields a score of 8.92, triggering a hot standby strategy: pre-deploying Spark compute container images on the edge, but delaying container instantiation until the actual peak traffic arrives.

[0109] In a refinery reactor monitoring scenario, this embodiment of the present invention uses A1 = 1 hour, a1 = 4 minutes to collect 15 pressure sensor data points, and B1 = 20 minutes, b1 = 2 minutes to collect 10 control command response latency data points. The environmental prediction model inputs the pressure fluctuation sequence into a four-layer fully connected network with a Swish activation function, outputting a score of 7.8. The business model analyzes the latency sequence using an RNN with an update gate structure, achieving a score of 6.3. When the total score of 6.9 reaches the first threshold, the backup Docker node is automatically dispatched to the 5G MEC server closest to the reactor, but the container instance is not activated, remaining in a preloaded state.

[0110] In a wind turbine fault warning scenario, this embodiment of the present invention sets A1 = 10 hours, a1 = 5 minutes to collect wind speed and gearbox vibration frequency; B1 = 2 hours, b1 = 2 minutes to collect SCADA system alarm logs. The environmental sequence is Z-score normalized and then fed into a multi-head attention module with Pre-ReLU, outputting a score of 8.7. The business model uses a BiLSTM to process log features, achieving a score of 7.1. Because the overall score of 8.1 exceeds 90% of the second threshold of 9.0, a tiered capacity expansion is triggered: three data analysis pods are first added to the local cluster, while GPU computing power reservation is requested from the regional data center.

[0111] The embodiment of the present invention uses an environment prediction model and a business prediction model to perform time series analysis on environment data and business data respectively, to achieve the functions of load change prediction and automatic container orchestration, thereby reducing service response delay.

[0112] The second embodiment of the present invention further discloses a method for orchestrating containers in the energy industry internet, comprising the following steps:

[0113] S1. Within a first preset time period A1, environmental data is collected at a first preset time interval a1 and normalized to form a plurality of sequence units to construct an environmental sequence;

[0114] S2. Within a second preset time period B1, business data is collected at a second preset time interval b1 and normalized to form multiple sequence units to construct a business sequence;

[0115] S3. Input the environmental sequence into the trained environmental prediction model and output the environmental impact score for the future time period Q;

[0116] S4. Input the business sequence into the trained business prediction model and output the business impact score for the future time period Q;

[0117] S5. Perform a weighted summation of the environmental impact score and the business impact score in the future time period Q to obtain a comprehensive impact score, and generate a container early warning orchestration strategy based on the comprehensive impact score.

[0118] In one embodiment of the present invention, the number of sequence units in the environment sequence is the ratio of the first preset time period A1 to the first preset time interval a1; the number of sequence units in the service sequence is the ratio of the second preset time period B1 to the second preset time interval b1.

[0119] This embodiment also provides a computer program product. When the computer program product runs on a computer, it enables the computer to execute the above-mentioned related steps to implement the energy industrial Internet container orchestration method provided in the above embodiment.

[0120] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in the embodiments of the present invention are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0121] Those skilled in the art will appreciate that the modules in the devices in the embodiments of the present invention can be adaptively changed and set in one or more devices different from the embodiments. The modules or units or components in the embodiments of the present invention can be combined into one module or unit or component, and in addition they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including corresponding claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner can be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including corresponding claims, abstracts and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0122] Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0123] It should be noted that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0124] In addition, each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. In particular, for embodiments such as devices and equipment, since they are basically similar to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The embodiments of the devices and equipment described above are merely schematic, wherein the modules, units, etc. described as separate components may or may not be physically separated, that is, they may be located in one place, or they may be distributed to multiple places, such as nodes in a system network. Specifically, some or all of the modules and units may be selected according to actual needs to achieve the purpose of the above-mentioned embodiment scheme. Those skilled in the art can understand and implement it without paying any creative work.

[0125] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0126] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0127] In addition, the terms "first" and "second" used in the embodiments of the present invention are only used for descriptive purposes and should not be understood as indicating or implying relative importance, or implicitly indicating the number of technical features indicated in this embodiment. Therefore, the features defined by the terms "first" and "second" in the embodiments of the present invention can explicitly or implicitly indicate that the embodiment includes at least one of such features. In the description of the present invention, the word "plurality" means at least two or two or more, such as two, three, four, etc., unless otherwise clearly and specifically defined in the embodiments.

[0128] In the embodiments of the present invention, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or apparatus. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or apparatus comprising the element. In addition, components, features, and elements with the same name in different embodiments of the present invention may have the same meaning or different meanings, and their specific meanings need to be determined by their explanation in the specific embodiment or further combined with the context of the specific embodiment.

[0129] Although embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention, and those of ordinary skill in the art may change, modify, replace, and modify the above embodiments within the scope of the present invention. Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the present invention. This application is intended to cover any variations, uses, or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art that are not disclosed in the present invention. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present invention are indicated by the claims below.

Claims

1. An energy industry Internet container orchestration security management system, characterized by: It includes an environment sequence building module, a business sequence building module, an environment impact prediction module, a business impact prediction module, and a container early warning orchestration module; The environment sequence construction module is used to collect environment data at a first preset time interval a1 within a first preset time period A1, and perform normalization processing to form multiple sequence units to construct an environment sequence; The service sequence construction module is used to collect service data according to the second preset time interval b1 within the second preset time period B1, and perform normalization processing to form multiple sequence units to construct a service sequence; The environmental impact prediction module is used to input the environmental sequence into the trained environmental prediction model and output the environmental impact score in the future time period Q; The business impact prediction module is used to input the business sequence into the trained business prediction model and output the business impact score in the future time period Q; The container early warning orchestration module is used to perform a weighted summation based on the environmental impact score and the business impact score in the future time period Q to obtain a comprehensive impact score, and generate a container early warning orchestration strategy based on the comprehensive impact score.

2. The energy industry internet container orchestration security management system according to claim 1 is characterized in that: The number of sequence units in the environment sequence is the ratio of the first preset time period A1 to the first preset time interval a1; the number of sequence units in the service sequence is the ratio of the second preset time period B1 to the second preset time interval b1.

3. The energy industry internet container orchestration security management system according to claim 1 is characterized in that: The environment prediction model is composed of M first hidden layers and a first classifier, wherein the number M of the first hidden layers is the same as the number of sequence units of the environment sequence; The mth sequence unit of the mth first hidden layer inputs the environment sequence and outputs the first update vector, 1≤m≤M; The first update vectors output by the M first hidden layers are input into the first classifier, and the environmental impact score in the future time period Q is represented by the classification space of the first classifier.

4. The energy industry internet container orchestration security management system according to claim 3 is characterized in that: The calculation formula of the first hidden layer includes: F m，1 =PReLU(h m ×W m，1 +b m，1 ); h m =Swish(X m ×W m，2 +h m-1 ×W m，3 +b m，2 ); Among them F m，1 represents the first update vector of the output of the mth first hidden layer, X m represents the mth sequence unit of the environment sequence input to the mth first hidden layer, h m and h m-1 They represent the hidden state vectors of the mth and m-1th first hidden layers respectively, and the dimension values ​​of h0 are all assigned to 0, W m，1 、W m，2 and W m，3 They represent the first weight parameter, the second weight parameter and the third weight parameter of the mth first hidden layer, respectively, and b m，1 and b m，2 They represent the first bias parameter and the second bias parameter of the mth first hidden layer, PReLU represents the PReLU activation function, and Swish represents the Swish activation function.

5. The energy industry internet container orchestration security management system according to claim 1 is characterized in that: The business prediction model is composed of N second hidden layers and second classifiers, wherein the number N of the second hidden layers is the same as the number of sequence units of the business sequence; The nth second hidden layer inputs the nth sequence unit of the business sequence and outputs the second update vector, 1≤n≤N; The second update vectors output by the N second hidden layers are input into the second classifier, and the business impact score in the future time period Q is represented by the classification space of the second classifier.

6. The energy industry internet container orchestration security management system according to claim 5 is characterized in that: The calculation formula of the second hidden layer includes: The reset gate r of the nth second hidden layer n The calculation formula is as follows: Where V n represents the nth sequence unit of the business sequence input to the nth second hidden layer, F n-1，2 represents the second update vector of the n-1th second hidden layer output, F 0，2 The dimension values ​​of are all assigned to 0. and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the reset gate of the nth second hidden layer, respectively. Sigmoid represents the Sigmoid activation function. The update gate z of the nth second hidden layer n The calculation formula is as follows: in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the update gate of the nth second hidden layer respectively; Candidate hidden states of the nth second hidden layer The calculation formula is as follows: in and They represent the first weight parameter, the second weight parameter, and the bias parameter corresponding to the candidate hidden state of the nth second hidden layer, respectively, and tanh represents the hyperbolic tangent function; The second update vector F output by the nth second hidden layer n，2 The calculation formula is as follows: Where ⊙ represents point-by-point multiplication.

7. The energy industry internet container orchestration security management system according to claim 1 is characterized in that: The environment prediction model and the business prediction model are both trained using preset training samples, and the sample labels of the training samples are obtained through manual evaluation and annotation; The environmental impact score output by the environmental prediction model ranges from 0 to 10; The business impact score output by the business prediction model ranges from 0 to 10.

8. The energy industry internet container orchestration security management system according to claim 1 is characterized in that: The weight coefficients corresponding to the environmental impact score and the business impact score obtained by weighted summation to obtain the comprehensive impact score are preset parameters; When the comprehensive score is less than the first score, there is no need to adjust the container; When the comprehensive score is between a first score and a second score, performing hot standby processing on the container in advance; the first score is less than the second score; When the comprehensive score is greater than the second score, the container is immediately expanded.

9. A container orchestration method for energy industry internet, characterized in that: The following steps are involved: In a first preset time period A1, environmental data is collected according to a first preset time interval a1, and normalized to form a plurality of sequence units to construct an environmental sequence; In a second preset time period B1, service data is collected according to the second preset time interval b1, and normalized to form a plurality of sequence units to construct a service sequence; Input the environmental sequence into the trained environmental prediction model and output the environmental impact score in the future time period Q; Input the business sequence into the trained business prediction model and output the business impact score in the future time period Q; A weighted sum of the environmental impact score and the business impact score in the future time period Q is performed to obtain a comprehensive impact score, and a container early warning orchestration strategy is generated based on the comprehensive impact score.

10. The method for orchestrating containers in the energy industry internet according to claim 9, characterized in that: The number of sequence units in the environment sequence is the ratio of the first preset time period A1 to the first preset time interval a1; the number of sequence units in the service sequence is the ratio of the second preset time period B1 to the second preset time interval b1.

Images