Method and device for determining abnormal service calling data, equipment and storage medium

Abstract

The invention discloses a method and device for determining abnormal service calling data, equipment and a storage medium. The method comprises the following steps: acquiring original historical service calling data; processing the original service call data by using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data in a target time period, the starting time in the target time period being later than the ending time in a historical time period corresponding to the original historical service call data; and determining abnormal service calling data in the target time period according to the predicted service calling data. According to the technical scheme, abnormal service calling is effectively monitored, and potential service calling problems are found and solved in time.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment and storage medium for determining abnormal service call data. Background Art

[0002] With the rapid construction of power systems and the development of digital transformation of project management business, the project management system of power grids is also booming.

[0003] As an important part of the project management system, the project management business platform focuses on the pain points, breakpoints and blind spots of project management. It can improve management efficiency and benefits internally, effectively connect with regulatory requirements externally, support project management standards and specifications, coordinate resource allocation, and reshape the digital architecture of project management.

[0004] However, the existing abnormal monitoring of the project management business middle platform mainly focuses on the response time and call volume of the middle platform services, lacking in-depth analysis of service call data such as front-end application usage, service support, and data changes. This makes it impossible to better evaluate the standardization and effectiveness of the middle platform service calls, and even affects the smooth development of project management business. Summary of the invention

[0005] The present invention provides a method, an apparatus, a device and a storage medium for determining abnormal service call data, so as to solve the problem of lack of effective identification of abnormal service call data.

[0006] In a first aspect, the present invention provides a method for determining abnormal service call data, comprising:

[0007] Get the original historical service call data;

[0008] Processing the original service call data using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data;

[0009] Abnormal service call data within the target time period is determined according to the predicted service call data.

[0010] In a second aspect, the present invention provides a device for determining abnormal service call data, comprising:

[0011] Data acquisition module, used to obtain original historical service call data;

[0012] A data prediction module, used for processing the original service call data by using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data;

[0013] The abnormal data identification module is used to determine the abnormal service call data within the target time period according to the predicted service call data.

[0014] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0015] at least one processor;

[0016] and a memory communicatively coupled to the at least one processor;

[0017] The memory stores a computer program executable by at least one processor, and the computer program is executed by at least one processor so that the at least one processor can execute the method for determining abnormal service call data of the first aspect.

[0018] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer instructions, which are used to enable a processor to implement the method for determining abnormal service call data of the first aspect when executed.

[0019] The present invention provides a solution for determining abnormal service call data, obtains original historical service call data, and uses a preset bidirectional long short-term memory network BiLSTM to process the original service call data to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data, and determines the abnormal service call data within the target time period based on the predicted service call data. By adopting the above technical solution, the service call data for future time periods is predicted by utilizing the selective memory characteristics of the long short-term memory network and the changing trend of historical data, and then the predicted service call data is used to identify abnormal service call data, effectively monitor abnormal service calls in different services in the project middle platform, and construct a more comprehensive, in-depth and intelligent monitoring method, better grasp the operating status of the project middle platform, and realize the timely discovery and resolution of potential service call problems.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 is a flowchart of a method for determining abnormal service call data provided according to Embodiment 1 of the present invention;

[0023] Figure 2 is a flowchart of a method for determining abnormal service call data provided according to Embodiment 2 of the present invention;

[0024] Figure 3 This is a result schematic diagram provided according to the second embodiment of the present invention;

[0025] Figure 4 It is a structural diagram of a device for determining abnormal service call data provided according to Embodiment 3 of the present invention;

[0026] Figure 5 It is a schematic diagram of the structure of an electronic device provided according to Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In the description of the present invention, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are an "or" relationship. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] Embodiment 1

[0030] Figure 1 A flowchart of a method for determining abnormal service call data is provided for embodiment 1 of the present invention. This embodiment is applicable to the case of determining abnormal service call data. The method can be executed by a device for determining abnormal service call data. The device for determining abnormal service call data can be implemented in the form of hardware and / or software. The device for determining abnormal service call data can be configured in an electronic device. The electronic device can be composed of two or more physical entities or one physical entity.

[0031] like Figure 1 As shown, the method for determining abnormal service call data provided by the first embodiment of the present invention specifically includes the following steps:

[0032] S101. Obtain original historical service call data.

[0033] S102. Use a preset bidirectional long short-term memory network BiLSTM to process the original service call data to obtain predicted service call data within a target time period, wherein the start time of the target time period is later than the end time of the historical time period corresponding to the original historical service call data.

[0034] In this embodiment, the bidirectional long short-term memory network BiLSTM can be used to process service exception data. As a sequence learning architecture, BiLSTM can selectively filter and forget certain information in the previous sequence to minimize its impact on subsequent sequences. After the original service call data is input into the preset bidirectional long short-term memory network BiLSTM, the network can predict the (predicted) service call data within the target time period. For example, by inputting the service call data two days before the current time into the preset BiLSTM, the predicted service call data one day before the current time can be output.

[0035] S103: Determine abnormal service call data within the target time period according to the predicted service call data.

[0036] In this embodiment, the predicted service call data that differs greatly from the actual service call data within the target time period may be determined as abnormal service call data within the target time period.

[0037] The method for determining abnormal service call data provided by the embodiment of the present invention obtains original historical service call data, and uses a preset bidirectional long short-term memory network BiLSTM to process the original service call data to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data, and the abnormal service call data within the target time period is determined based on the predicted service call data. The technical solution of the embodiment of the present invention uses the selective memory characteristics of the long short-term memory network and the changing trend of historical data to predict the service call data of the future time period, and then uses the predicted service call data to realize the identification of abnormal service call data, effectively monitor the abnormal service calls in different services of the project middle platform, and construct a more comprehensive, in-depth and intelligent monitoring method, better grasp the operating status of the project middle platform, and realize the timely discovery and resolution of potential service call problems.

[0038] Optionally, the using a preset bidirectional long short-term memory network BiLSTM to process the original service call data to obtain predicted service call data within a target time period includes: using the input layer in the preset bidirectional long short-term memory network BiLSTM to standardize the original service call data to obtain multiple input samples; using the hidden layer in the BiLSTM to process the multiple input samples to obtain initial prediction data, wherein the hidden layer includes a two-layer LSTM network and a Dense fully connected layer; using the output layer in the BiLSTM to de-standardize the initial prediction data to obtain the predicted service call data within the target time period.

[0039] Specifically, the preset bidirectional long short-term memory network BiLSTM of this solution includes a two-layer LSTM network and a Dense fully connected layer, which can use the characteristics of LSTM to process sequence data, such as time series analysis, natural language processing, or any task involving dependencies between previous and subsequent data points. The stacked two LSTM layers can enhance the learning ability of the model, enabling BiLSTM to capture more complex data patterns and time dependencies.

[0040] For example, suppose that the original service call data in the historical time period t is D=X1, X2, ..., Xm*n. The service call data can be standardized using the input layer, and m data are set as 1 group, with a total of n groups. Each group of data corresponds to an input sample xi, thereby obtaining n input samples. Then, the input sample is processed using the hidden layer in BiLSTM to obtain the initial prediction data. Finally, the initial prediction data is iterated and de-standardized using the output layer in BiLSTM to obtain the predicted service call data with the same dimension as the actual service call data.

[0041] Optionally, using the hidden layer in the BiLSTM to process the multiple input samples to obtain initial prediction data includes: using the first layer LSTM network in the hidden layer of the BiLSTM to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships, and performing forgetting and memory operations on the sample data to obtain target features; using the second layer LSTM network in the hidden layer of the BiLSTM to convert the target features into a multivariate having a linear relationship with the prediction service call data; using the Dense fully connected layer in the hidden layer of the BiLSTM to process the multivariate to obtain initial prediction data.

[0042] Specifically, the first layer of LSTM network can implicitly project the input xi into a high-dimensional space through its specific network structure to obtain sample data containing the association relationship of service call data, and make judgments through the unique gate structure in the LSTM network to decide whether to remember or forget certain features to obtain the target features. The second layer of LSTM network refines the target features into multiple variables that have a linear relationship with the pre-recovered data. Finally, the data prediction problem is simplified to a linear regression problem through the fully connected layer Dense. By processing multiple variables, the initial prediction data can be obtained.

[0043] Embodiment 2

[0044] Figure 2 A flowchart of a method for determining abnormal service call data provided in the second embodiment of the present invention. The technical solution of the embodiment of the present invention is further optimized on the basis of the above-mentioned optional technical solutions, and a specific method for determining abnormal service call data is given.

[0045] Optionally, the hidden layer also includes a feature attention layer and a timing attention layer; wherein, before using the first layer LSTM network in the hidden layer of the BiLSTM to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships, it also includes: using the feature attention layer in the hidden layer of the BiLSTM to process the multiple input samples to obtain an input feature sequence; wherein, using the first layer LSTM network in the hidden layer of the BiLSTM to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships includes: using the first layer LSTM network in the hidden layer of the BiLSTM to implicitly project the input feature sequence into a high-dimensional space to obtain sample data containing association relationships; after using the Dense fully connected layer in the hidden layer of the BiLSTM to process the multiple variables to obtain initial prediction data, it also includes: using the timing attention layer in the hidden layer of the BiLSTM to process the initial prediction data to obtain initial prediction data containing time series information. The advantage of this setting is that by introducing the attention layer in BiLSTM, the importance of input information can be measured, and different weights can be assigned to different input information according to their importance, which improves the judgment ability of LSTM without affecting the storage and computational complexity of the original LSTM model, greatly improving the efficiency of the training model and the accuracy of the prediction results.

[0046] Optionally, determining the abnormal service call data within the target time period according to the predicted service call data includes: determining the abnormal service call data within the target time period according to the deviation between the predicted service call data and the actual service call data within the target time period. The advantage of such a setting is that by comparing the gap between the predicted service call data and the actual service call data, the abnormal service call data can be quickly and accurately determined.

[0047] like Figure 2 As shown, a method for determining abnormal service call data provided by Embodiment 2 of the present invention specifically includes the following steps:

[0048] S201. Obtain original historical service call data.

[0049] S202: Standardize the original service call data using the input layer in a preset bidirectional long short-term memory network BiLSTM to obtain multiple input samples.

[0050] S203. Use the feature attention layer in the hidden layer of the BiLSTM to process the multiple input samples to obtain an input feature sequence.

[0051] Specifically, the attention mechanism can enhance the performance of the LSTM model. In order to overcome the limitations of the traditional LSTM encoding process, the attention mechanism can be introduced into the LSTM to better preserve the intermediate state of the encoder, amplify key information and prevent the traditional LSTM network from forgetting such information. The attention mechanism greatly improves the speed of information reception and processing by focusing on certain information and ignoring irrelevant information while amplifying useful information. The attention mechanism can dynamically set the weight of the input features during the calculation process, and reflect the correlation differences between different features and output results through the change of attention, so as to obtain better prediction results. Information with larger weights will focus attention, which can directly affect the accuracy of the model.

[0052] Among them, the feature attention mechanism can assign attention weights to the input features of the LSTM network and dynamically update them to determine the connection strength between the input features and the target features, focusing on key features and minimizing or ignoring irrelevant features, thereby improving the accuracy of obtaining decisive input features in the LSTM network. In the feature attention layer, the attention mechanism can dynamically extract feature information with important correlations for multiple input samples through weight parameters, thereby obtaining the input feature sequence of the LSTM network.

[0053] S204, using the first layer LSTM network in the hidden layer of the BiLSTM, implicitly projecting the input feature sequence into a high-dimensional space to obtain sample data containing association relationships, and performing forgetting and memorizing operations on the sample data to obtain target features.

[0054] S205. Using the second LSTM network in the hidden layer of the BiLSTM, the target feature is converted into a multivariate having a linear relationship with the predicted service call data.

[0055] S206. Using the Dense fully connected layer in the hidden layer of the BiLSTM, the multivariate is processed to obtain initial prediction data.

[0056] Specifically, the LSTM network and the Dense fully connected layer can extract the time correlation information hidden in the historical service call data to obtain the hidden layer state at each historical moment, that is, the initial prediction data.

[0057] S207. Using the temporal attention layer in the hidden layer of the BiLSTM, the initial prediction data is processed to obtain initial prediction data containing time series information.

[0058] Specifically, the temporal attention mechanism can assign attention weights to important time information to help BiLSTM determine the degree of correlation between the current data and the output. At the same time, the above steps can obtain the relevance of historical service call data on a time scale and set weight coefficients for historical service call data, where key data has a greater weight so that its impact on the output can be fully utilized. The temporal attention layer can assign attention weights to the time information in the initial prediction data, so that the initial prediction data containing time series information can be obtained.

[0059] S208. De-normalize the initial prediction data using the output layer in the BiLSTM to obtain predicted service call data within a target time period.

[0060] Specifically, the output layer in BiLSTM can perform de-normalization processing such as superposition and reconstruction on the initial prediction data to obtain the predicted service call data within the target time period.

[0061] S209: Determine abnormal service call data within the target time period according to the deviation between the predicted service call data and the actual service call data within the target time period.

[0062] Specifically, the service call data within the target time period corresponding to the larger deviation may be determined as abnormal service call data.

[0063] Optionally, determining abnormal service call data within the target time period based on the deviation between the predicted service call data and the actual service call data within the target time period includes: determining the deviation between the predicted service call data and the actual service call data within the target time period; determining a target deviation greater than a preset threshold and a time corresponding to the target deviation, and determining the actual service call data corresponding to the corresponding time as abnormal service call data.

[0064] For example, Figure 3 The original historical service call data may be service call data from January 2023 to June 2024, and some abnormal points are artificially set for comparison with the existing clustering method and rule method. Figure 3The blue line in the middle represents the service call data within the target time period, the red circles 1 and 2 represent the abnormal service call data identified by this solution, and the red circle 3 represents the abnormal service call data identified by the traditional method (including the traditional clustering anomaly detection method and the traditional rule-based anomaly detection method) and this solution. When the deviation is small, such as less than 10%, only this solution can effectively identify abnormal data, and this solution has good adaptability. The solution can also be generalized using multiple sets of independent service call data. The test results show that this solution performs well in monitoring abnormal service call data, with the highest monitoring accuracy reaching 99.8% and the average monitoring accuracy reaching 98.77%.

[0065] The method for determining abnormal service call data provided by an embodiment of the present invention realizes the measurement of the importance of input information by introducing an attention layer in BiLSTM, and can assign different weights to different input information according to the importance, thereby improving the judgment ability of LSTM without affecting the storage and computational complexity of the original LSTM model. It greatly improves the efficiency of the training model and the accuracy of the prediction results, and finally, by comparing the gap between the predicted service call data and the actual service call data, the abnormal service call data can be determined quickly and accurately.

[0066] Embodiment 3

[0067] Figure 4 This is a schematic diagram of a structure of a device for determining abnormal service call data provided by Embodiment 3 of the present invention. Figure 4 As shown, the device includes: a data acquisition module 301, a data prediction module 302 and an abnormal data identification module 303, wherein:

[0068] Data acquisition module, used to obtain original historical service call data;

[0069] A data prediction module, used for processing the original service call data by using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data;

[0070] The abnormal data identification module is used to determine the abnormal service call data within the target time period according to the predicted service call data.

[0071] The device for determining abnormal service call data provided by the embodiment of the present invention utilizes the selective memory characteristics of the long short-term memory network and the changing trend of historical data to predict the service call data of the future time period, and then utilizes the predicted service call data to realize the identification of abnormal service call data, effectively monitors the abnormal service calls in different services of the project middle platform, constructs a more comprehensive, in-depth and intelligent monitoring method, better grasps the operating status of the project middle platform, and realizes the timely discovery and resolution of potential service call problems.

[0072] Optionally, the data prediction module includes:

[0073] An input layer processing unit, used for performing standardization processing on the original service call data by using the input layer in a preset bidirectional long short-term memory network BiLSTM to obtain a plurality of input samples;

[0074] A hidden layer processing unit, used for processing the multiple input samples using the hidden layer in the BiLSTM to obtain initial prediction data, wherein the hidden layer includes a double-layer LSTM network and a Dense fully connected layer;

[0075] The output layer processing unit is used to use the output layer in the BiLSTM to perform de-standardization processing on the initial prediction data to obtain the prediction service call data within the target time period.

[0076] Furthermore, the use of the hidden layer in the BiLSTM to process the multiple input samples to obtain initial prediction data includes: using the first layer LSTM network in the hidden layer of the BiLSTM to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships, and performing forgetting and memory operations on the sample data to obtain target features; using the second layer LSTM network in the hidden layer of the BiLSTM to convert the target features into a multivariate having a linear relationship with the prediction service call data; and using the Dense fully connected layer in the hidden layer of the BiLSTM to process the multivariate to obtain initial prediction data.

[0077] Optionally, the hidden layer also includes a feature attention layer and a temporal attention layer.

[0078] Optionally, the data prediction module also includes:

[0079] The feature sequence determination unit is used to process the multiple input samples using the feature attention layer in the hidden layer of the BiLSTM to obtain an input feature sequence before implicitly projecting the multiple input samples into a high-dimensional space using the first layer LSTM network in the hidden layer of the BiLSTM to obtain sample data containing association relationships.

[0080] Optionally, the first LSTM network layer in the hidden layer of the BiLSTM is used to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships, including: the first LSTM network layer in the hidden layer of the BiLSTM is used to implicitly project the input feature sequence into a high-dimensional space to obtain sample data containing association relationships.

[0081] Optionally, the data prediction module also includes:

[0082] The initial prediction data determination unit is used to process the multiple variables using the Dense fully connected layer in the hidden layer of the BiLSTM to obtain the initial prediction data, and then use the temporal attention layer in the hidden layer of the BiLSTM to process the initial prediction data to obtain the initial prediction data containing time series information.

[0083] Optionally, the abnormal data identification module is specifically used to determine the abnormal service call data within the target time period according to the deviation between the predicted service call data and the actual service call data within the target time period.

[0084] Furthermore, determining abnormal service call data within the target time period based on the deviation between the predicted service call data and the actual service call data within the target time period includes: determining the deviation between the predicted service call data and the actual service call data within the target time period; determining a target deviation greater than a preset threshold and a time corresponding to the target deviation, and determining the actual service call data corresponding to the corresponding time as abnormal service call data.

[0085] Optionally, the original historical service call data includes original service call data of power system projects within a historical time period.

[0086] The apparatus for determining abnormal service call data provided by the embodiment of the present invention can execute the method for determining abnormal service call data provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0087] Embodiment 4

[0088] Figure 5A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0089] like Figure 5 As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0090] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0091] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 41 executes the various methods and processes described above, such as determining an abnormal service call data method.

[0092] In some embodiments, the method for determining abnormal service call data may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the method for determining abnormal service call data described above may be performed. Alternatively, in other embodiments, the processor 41 may be configured to execute the method for determining abnormal service call data in any other appropriate manner (e.g., by means of firmware).

[0093] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0095] The computer device provided above can be used to execute the method for determining abnormal service call data provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0096] Embodiment 5

[0097] In the context of the present invention, a computer-readable storage medium may be a tangible medium, wherein the computer-executable instructions, when executed by a computer processor, are used to perform a method for determining abnormal service call data, the method comprising:

[0098] Get the original historical service call data;

[0099] Processing the original service call data using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data;

[0100] Abnormal service call data within the target time period is determined according to the predicted service call data.

[0101] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by an instruction execution system, device or equipment or used with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0102] The computer device provided above can be used to execute the method for determining abnormal service call data provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0103] It is worth noting that in the above-mentioned embodiment of the device for determining abnormal service call data, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0104] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for determining abnormal service call data, characterized in that: include: Get the original historical service call data; Processing the original service call data using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data; Abnormal service call data within the target time period is determined according to the predicted service call data.

2. The method according to claim 1, characterized in that The using a preset bidirectional long short-term memory network BiLSTM to process the original service call data to obtain predicted service call data within a target time period includes: The original service call data is standardized using an input layer in a preset bidirectional long short-term memory network BiLSTM to obtain a plurality of input samples; Processing the multiple input samples using a hidden layer in the BiLSTM to obtain initial prediction data, wherein the hidden layer includes a double-layer LSTM network and a Dense fully connected layer; The initial prediction data is de-normalized using the output layer in the BiLSTM to obtain the predicted service call data within the target time period.

3. The method according to claim 2, characterized in that The using the hidden layer in the BiLSTM to process the multiple input samples to obtain initial prediction data includes: Using the first LSTM network in the hidden layer of the BiLSTM, the multiple input samples are implicitly projected into a high-dimensional space to obtain sample data containing association relationships, and forgetting and memorizing operations are performed on the sample data to obtain target features; Using a second LSTM network in the hidden layer of the BiLSTM, the target feature is converted into a multivariate having a linear relationship with the predicted service call data; The multivariate is processed using the Dense fully connected layer in the hidden layer of the BiLSTM to obtain initial prediction data.

4. The method according to claim 3, characterized in that The hidden layer also includes a feature attention layer and a temporal attention layer; wherein, before the first layer LSTM network in the hidden layer of the BiLSTM is used to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships, it also includes: Processing the plurality of input samples using a feature attention layer in a hidden layer of the BiLSTM to obtain an input feature sequence; The method of using the first LSTM network in the hidden layer of the BiLSTM to implicitly project the multiple input samples into a high-dimensional space to obtain sample data containing association relationships includes: Using the first LSTM network in the hidden layer of the BiLSTM, the input feature sequence is implicitly projected into a high-dimensional space to obtain sample data containing an association relationship; After the Dense fully connected layer in the hidden layer of the BiLSTM is used to process the multivariate to obtain initial prediction data, the method further includes: The initial prediction data is processed using the temporal attention layer in the hidden layer of the BiLSTM to obtain initial prediction data containing time series information.

5. The method according to claim 1, characterized in that The determining the abnormal service call data within the target time period according to the predicted service call data includes: Abnormal service call data within the target time period is determined according to a deviation between the predicted service call data and the actual service call data within the target time period.

6. The method according to claim 5, characterized in that The determining, according to the deviation between the predicted service call data and the actual service call data within the target time period, the abnormal service call data within the target time period comprises: Determining a deviation between the predicted service call data and actual service call data within the target time period; A target deviation greater than a preset threshold and a time corresponding to the target deviation are determined, and actual service call data corresponding to the corresponding time are determined as abnormal service call data.

7. The method according to claim 1, characterized in that The original historical service call data includes original service call data of the power system project within a historical time period.

8. A device for determining abnormal service call data, characterized in that: include: Data acquisition module, used to obtain original historical service call data; A data prediction module, used for processing the original service call data by using a preset bidirectional long short-term memory network BiLSTM to obtain predicted service call data within a target time period, wherein the start time within the target time period is later than the end time within the historical time period corresponding to the original historical service call data; The abnormal data identification module is used to determine the abnormal service call data within the target time period according to the predicted service call data.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for determining abnormal service call data according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining abnormal service call data according to any one of claims 1 to 7 when executed.

Images