CN117850258A - Abnormality detection method and device for Internet of things equipment, storage medium and electronic equipment - Google Patents

Abstract

The application discloses an anomaly detection method and device for Internet of things equipment and electronic equipment. The method comprises the following steps: collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment has abnormality, wherein the target model is trained by a plurality of groups of training samples, and each group of training samples comprises a group of historical signal characteristics and a historical detection result of a historical detection signal. Through the method and the device, the problem of low detection efficiency of the abnormal state of the Internet of things equipment in the related technology is solved.

Description

Abnormality detection method and device for Internet of things equipment, storage medium and electronic equipment

Technical Field

The application relates to the field of internet of things equipment, in particular to an abnormality detection method and device for internet of things equipment and electronic equipment.

Background

Along with development of science and technology, intelligent home has been popularized in the home of each user, and the internet of things equipment in the intelligent home is various, can lead to various abnormal conditions to appear in the equipment based on various reasons in the in-service use of internet of things equipment, and in the related art, the abnormal detection of internet of things equipment is usually after the internet of things equipment breaks down a period of time, just can discover the trouble that the internet of things equipment appears by professional maintainer to detect the internet of things equipment, and is lower to the abnormal detection efficiency of internet of things equipment.

Aiming at the problem of low detection efficiency of abnormal states of Internet of things equipment in the related technology, no effective solution is proposed at present.

Disclosure of Invention

The main object of the present application is to provide an anomaly detection method and apparatus for an internet of things device, and an electronic device, so as to solve the problem of low anomaly detection efficiency of the internet of things device in the related art.

In order to achieve the above object, according to one aspect of the present application, there is provided an anomaly detection method for an internet of things device. The method comprises the following steps: collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment has abnormality, wherein the target model is trained by a plurality of groups of training samples, and each group of training samples comprises a group of historical signal characteristics and a historical detection result of a historical detection signal.

Optionally, extracting signal features from the target signal, to obtain a set of signal features to be detected includes: analyzing the target signal according to the time sequence to obtain a periodic waveform of the target signal; extracting statistical features, energy features and frequency domain features of the target signal from the periodic waveform; determining a set of signal characteristics to be detected of the target signal, wherein the set of signal characteristics to be detected comprises at least one of: statistical features, energy features, and frequency domain features.

Optionally, after obtaining the detection result for characterizing whether the target internet of things device has an abnormality, the method further includes: under the condition that the detection result represents that the target Internet of things equipment is abnormal, sending out prompt information, wherein the prompt information is used for prompting a user that the target Internet of things equipment needs to be maintained; and determining the abnormal state grade of the target Internet of things equipment, and executing preset protection measures on the target Internet of things equipment based on the abnormal state grade.

Optionally, determining the abnormal state level of the target internet of things device, and executing the preset protection measure on the target internet of things device based on the abnormal state level includes: extracting an abnormal state evaluation value of the target internet of things equipment from the detection result, and determining a target abnormal state evaluation range to which the abnormal state evaluation value belongs, wherein the target internet of things equipment is preset with a plurality of abnormal state grades, and each abnormal state grade corresponds to one abnormal state evaluation range; determining a target abnormal state grade corresponding to the target abnormal state evaluation range, and executing a preset protection measure corresponding to the target abnormal state grade on the target internet of things equipment, wherein the preset protection measure at least comprises one of the following steps: pausing the function, shutdown, power outage and networking diagnostics of the target internet of things device.

Optionally, collecting the detection signal of the target sensor on the target internet of things device includes: determining the equipment type of the target Internet of things equipment, and determining a target sensor required for collecting detection signals of the target Internet of things equipment from a preset corresponding relation set through the equipment type, wherein the preset corresponding relation set comprises a plurality of corresponding relations, and each corresponding relation comprises the Internet of things equipment of one equipment type and at least one sensor for collecting the detection signals of the Internet of things equipment.

Alternatively, the object model is obtained by: acquiring a plurality of history detection signals, and extracting a group of history signal characteristics from each history detection signal; determining a history detection result of each history detection signal, and taking a group of history signal characteristics and the history detection result of each history detection signal as a group of training samples to obtain a plurality of groups of training samples; training a preset classification model based on a plurality of groups of training samples to obtain a target model, wherein the preset classification model at least comprises one of the following: logistic regression models, decision tree models, and random forest models.

Optionally, training a preset classification model based on multiple sets of training samples, and obtaining the target model includes: dividing a plurality of groups of training samples into a training set, a verification set and a test set according to a preset proportion; determining N preset classification models, and training the preset classification models through a training set for each preset classification model to obtain an initial model, wherein N is a positive integer; adjusting model parameters of the initial model, and determining a generalization capability evaluation value of the initial model after each adjustment of the model parameters through a verification set to obtain a group of generalization capability evaluation values of the initial model; determining an initial model corresponding to a maximum generalization capability evaluation value in a set of generalization capability evaluation values as a pending model; determining performance evaluation values of the undetermined models corresponding to each preset classification model through the test set to obtain a group of performance evaluation values of N undetermined models; and determining a pending model corresponding to the maximum performance evaluation value in the group of performance evaluation values as a target model.

Optionally, in the case that there is a new internet of things device, the method further includes: acquiring detection signals of a target sensor on newly-added internet of things equipment in an abnormal state, obtaining abnormal detection signals, and extracting a group of abnormal detection signal characteristics through the abnormal detection signals; collecting detection signals of a target sensor on newly-added internet of things equipment in a normal state, obtaining normal detection signals, and extracting a group of normal detection signal characteristics through the normal detection signals; determining a group of abnormal detection signal characteristics and abnormal detection signals as a group of newly added training samples, and determining a group of normal detection signal characteristics and normal detection signals as a group of newly added training samples; combining the newly added training samples with a plurality of groups of training samples to obtain a plurality of groups of updated training samples; and training the target model based on the updated multiple groups of training samples to obtain an updated target model.

In order to achieve the above object, according to another aspect of the present application, there is provided an abnormality detection apparatus for an internet of things device. The device comprises: the system comprises an acquisition unit, a detection unit and a control unit, wherein the acquisition unit is used for acquiring detection signals of a target sensor on target internet of things equipment, and the target sensor at least comprises one of the following components: a power sensor, a voltage sensor, a current sensor and a temperature sensor; the processing unit is used for carrying out noise reduction processing on the detection signals to obtain target signals, extracting signal characteristics from the target signals and obtaining a group of signal characteristics to be detected; the input unit is used for inputting a group of signal characteristics to be detected into the target model to obtain a detection result used for representing whether the target internet of things equipment has an abnormality, wherein the target model is obtained by training a plurality of groups of training samples, and each group of training samples comprises a group of historical signal characteristics and a group of historical detection results of the historical detection signals.

Through the application, the following steps are adopted: collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment is abnormal, wherein the target model is trained by a plurality of groups of training samples, each group of training samples comprises a group of historical signal characteristics of historical detection signals and the historical detection result, and the problem that the abnormal state detection efficiency of the Internet of things equipment in the related technology is low is solved. And extracting signal characteristics to be detected through the acquired detection signals of the Internet of things equipment, and inputting the signal characteristics to be detected into a target model to obtain a detection result of whether the Internet of things equipment is abnormal or not, so that the effect of improving the abnormal state detection efficiency of the Internet of things equipment is achieved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application, illustrate and explain the application and are not to be construed as limiting the application. In the drawings:

Fig. 1 is a hardware block diagram of a mobile terminal according to an anomaly detection method of an internet of things device according to an embodiment of the present application;

fig. 2 is a flowchart of an anomaly detection method of an internet of things device according to an embodiment of the present application;

fig. 3 is a schematic diagram of an abnormality detection apparatus of an internet of things device according to an embodiment of the present application;

fig. 4 is a schematic diagram of an electronic device provided according to an embodiment of the present application.

Detailed Description

It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

In order to make the present application solution better understood by those skilled in the art, the following description will be made in detail and with reference to the accompanying drawings in the embodiments of the present application, it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the present application described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, etc.) and the data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by the user or sufficiently authorized by each party.

The method embodiments provided in the embodiments of the present invention may be performed in a mobile terminal, a computer terminal or similar computing device. Taking the operation on a mobile terminal as an example, fig. 1 is a block diagram of a hardware structure of the mobile terminal according to the abnormality detection method of the internet of things device provided in the embodiment of the present application. As shown in fig. 1, a mobile terminal may include one or more (only one is shown in fig. 1) processors 102 (the processor 102 may include, but is not limited to, a microprocessor MCU or a processing device such as a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 for communication functions and an input-output device 108. It will be appreciated by those skilled in the art that the structure shown in fig. 1 is merely illustrative and not limiting of the structure of the mobile terminal described above. For example, the mobile terminal may also include more or fewer components than shown in fig. 1, or have a different configuration than shown in fig. 1.

The memory 104 may be used to store a computer program, for example, a software program of application software and a module, such as a computer program corresponding to an abnormality detection method of an internet of things device in an embodiment of the present invention, and the processor 102 executes the computer program stored in the memory 104, thereby performing various functional applications and data processing, that is, implementing the above-mentioned method. Memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include memory remotely located relative to the processor 102, which may be connected to the mobile terminal via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or transmit data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, simply referred to as NIC) that can connect to other network devices through a base station to communicate with the internet. In one example, the transmission device 106 may be a Radio Frequency (RF) module, which is configured to communicate with the internet wirelessly.

The present invention is described below in connection with preferred implementation steps, and fig. 2 is a flowchart of an abnormality detection method for an internet of things device according to an embodiment of the present application, as shown in fig. 2, where the method includes the following steps:

step S201, collecting a detection signal of a target sensor on the target internet of things device, where the target sensor at least includes one of the following: power sensors, voltage sensors, current sensors, and temperature sensors.

Specifically, the target sensor is integrated in the target internet of things equipment, the output signal of the internet of things equipment is monitored in real time based on the target sensor, and the abnormal state of the target internet of things equipment is detected by collecting the detection signal in the target sensor.

For example, the target sensor may be a power sensor, through which the power consumption of the target internet of things device is collected and the output power waveform is recorded. Whether the target internet of things equipment is abnormal or not is detected through the power waveform, and the collected detection signals comprise the power waveforms of the target internet of things equipment in different operation modes so as to ensure that the normal working state of the target internet of things equipment in various modes is covered.

Step S202, noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected.

Specifically, it is difficult to directly extract the signal characteristics from the detection signal acquired by the target sensor, and thus it is necessary to perform noise reduction processing on the detection signal to obtain a target signal that can be used for extracting the signal characteristics, for example, to remove high-frequency noise in the detection signal by using a low-pass filter to obtain the target signal. The periodic pattern and trend of the target signal are identified by time series analysis of the target signal. Extracting signal characteristics from a periodic mode of a target signal, wherein a group of signal characteristics to be detected can comprise statistical characteristics such as mean, standard deviation, skewness and kurtosis; energy characteristics such as total power consumption and frequency domain characteristics such as main frequency components, energy distribution, etc.

Step S203, inputting a group of signal features to be detected into a target model to obtain a detection result for representing whether the target Internet of things equipment has an abnormality, wherein the target model is trained by a plurality of groups of training samples, and each group of training samples comprises a group of historical signal features and a group of historical detection results of the historical detection signals.

Specifically, the target model may be a classification model based on machine learning, such as a random forest model, a decision tree model, and the like, by collecting historical detection signals of the target internet of things equipment in a normal working state and in an abnormal working state, taking a set of historical signal characteristics and historical detection results based on the historical detection signals as training samples, training out a target model for automatically identifying whether the target internet of things equipment is abnormal through a plurality of sets of training samples, and then determining whether the target internet of things equipment is abnormal through the target model.

According to the anomaly detection method for the Internet of things equipment, detection signals of the target sensor on the target Internet of things equipment are collected, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment is abnormal, wherein the target model is trained by a plurality of groups of training samples, each group of training samples comprises a group of historical signal characteristics of historical detection signals and the historical detection result, and the problem that the abnormal state detection efficiency of the Internet of things equipment in the related technology is low is solved. And extracting signal characteristics to be detected through the acquired detection signals of the Internet of things equipment, and inputting the signal characteristics to be detected into a target model to obtain a detection result of whether the Internet of things equipment is abnormal or not, so that the effect of improving the abnormal state detection efficiency of the Internet of things equipment is achieved.

In order to efficiently identify whether an abnormality occurs in a target internet of things device, a set of signal features to be detected needs to be extracted from a target signal, optionally, in the abnormality detection method of the internet of things device provided by the embodiment of the present application, extracting the signal features from the target signal, and obtaining the set of signal features to be detected includes: analyzing the target signal according to the time sequence to obtain a periodic waveform of the target signal; extracting statistical features, energy features and frequency domain features of the target signal from the periodic waveform; determining a set of signal characteristics to be detected of the target signal, wherein the set of signal characteristics to be detected comprises at least one of: statistical features, energy features, and frequency domain features.

Specifically, statistical features such as mean, standard deviation, skewness, kurtosis; the energy characteristics such as total power consumption and frequency domain characteristics such as main frequency components, energy distribution and the like are analyzed according to time sequence to extract the periodic waveform of the target signal, and the characteristics which are most helpful to classification are selected from the periodic waveforms. And extracting the statistical characteristics, the energy characteristics and the frequency domain characteristics of the target signal through measures such as characteristic construction, characteristic scaling and the like. The detection feature for identifying whether the target internet of things device has an abnormality may include at least one of the following features: statistical features, energy features, and frequency domain features. The abnormality detection method based on the signal characteristics to be detected in the embodiment utilizes the power usage mode of the internet of things equipment to predict and identify potential faults. The method can be carried out under the condition of not affecting the normal operation of the equipment, and can early warn before the occurrence of the problem, thereby reducing the downtime and maintenance cost of the equipment.

It should be noted that feature construction is the creation of new features by computing or combining the original features. Feature scaling is the normalization or normalization of features to avoid the influence of large-scale features on model results.

By timely executing preset protection measures on the target internet of things equipment with the abnormality, the maintenance efficiency of the target internet of things equipment is improved, and optionally, in the abnormality detection method for the internet of things equipment provided by the embodiment of the application, after the detection result for representing whether the target internet of things equipment has the abnormality is obtained, the method further comprises the following steps: under the condition that the detection result represents that the target Internet of things equipment is abnormal, sending out prompt information, wherein the prompt information is used for prompting a user that the target Internet of things equipment needs to be maintained; and determining the abnormal state grade of the target Internet of things equipment, and executing preset protection measures on the target Internet of things equipment based on the abnormal state grade.

Specifically, when abnormality of the target internet of things equipment is detected, prompt information is automatically sent to an administrator or a maintenance team. Meanwhile, according to the abnormal state level of the target internet of things equipment, different preset protection measures are adopted to protect the target internet of things equipment, and the preset protection measures are used for suspending the functions, stopping, powering off, networking diagnosis and the like of the target internet of things equipment.

Optionally, in the anomaly detection method of the internet of things device provided in the embodiment of the present application, determining an anomaly state level of the target internet of things device, and executing a preset protection measure on the target internet of things device based on the anomaly state level includes: extracting an abnormal state evaluation value of the target internet of things equipment from the detection result, and determining a target abnormal state evaluation range to which the abnormal state evaluation value belongs, wherein the target internet of things equipment is preset with a plurality of abnormal state grades, and each abnormal state grade corresponds to one abnormal state evaluation range; determining a target abnormal state grade corresponding to the target abnormal state evaluation range, and executing a preset protection measure corresponding to the target abnormal state grade on the target internet of things equipment, wherein the preset protection measure at least comprises one of the following steps: pausing the function, shutdown, power outage and networking diagnostics of the target internet of things device.

Specifically, the normal state of the target internet of things device is the behavior or performance of the device or system when operating as intended. The abnormal state indicates that the behavior or performance deviating from the normal performance is presented, which indicates that the target internet of things equipment may have a problem or a fault. The detection result comprises an abnormal state evaluation value of the target internet of things equipment, a target abnormal state grade of the target internet of things equipment is determined through an abnormal state evaluation range which is manually set in advance, and further preset protection measures corresponding to the target abnormal state grade are automatically executed.

Optionally, in the anomaly detection method of the internet of things device provided in the embodiment of the present application, collecting a detection signal of a target sensor on a target internet of things device includes: determining the equipment type of the target Internet of things equipment, and determining a target sensor required for collecting detection signals of the target Internet of things equipment from a preset corresponding relation set through the equipment type, wherein the preset corresponding relation set comprises a plurality of corresponding relations, and each corresponding relation comprises the Internet of things equipment of one equipment type and at least one sensor for collecting the detection signals of the Internet of things equipment.

Specifically, because abnormal states of different types of internet of things equipment are different, detection signals required to be collected are also different, so that sensors applied to the different types of internet of things equipment are also different, and corresponding sensors for collecting the detection signals of the internet of things equipment are preset in advance for the different types of internet of things equipment by setting a preset corresponding relation set, for example, the A internet of things equipment needs a temperature sensor to collect the detection signals, the B internet of things equipment needs a voltage sensor and a current sensor to collect the detection signals, and the C internet of things equipment needs a power sensor to collect the detection signals. And determining a target sensor from a preset corresponding relation set according to the equipment type of the target Internet of things equipment, and further acquiring a detection signal of the target Internet of things equipment based on the target sensor.

In order to efficiently identify the abnormal situation of the target internet of things device, a target model needs to be trained, optionally, in the abnormality detection method of the internet of things device provided in the embodiment of the present application, the target model is obtained by the following manner: acquiring a plurality of history detection signals, and extracting a group of history signal characteristics from each history detection signal; determining a history detection result of each history detection signal, and taking a group of history signal characteristics and the history detection result of each history detection signal as a group of training samples to obtain a plurality of groups of training samples; training a preset classification model based on a plurality of groups of training samples to obtain a target model, wherein the preset classification model at least comprises one of the following: logistic regression models, decision tree models, and random forest models.

Specifically, a historical detection signal is collected, and a preset classification model is trained according to the historical signal characteristics and the historical detection result of the historical detection signal as training samples, so that a target model is obtained. It should be noted that, the training sample includes detection signals in the normal working state and the abnormal working state of the target internet of things device. According to the method, the device and the system, the cause of the fault of the target Internet of things device is found out more quickly and clearly through training the target model. The user can solve the equipment abnormality in advance.

Optionally, in the anomaly detection method for the internet of things device provided in the embodiment of the present application, training a preset classification model based on multiple sets of training samples, obtaining the target model includes: dividing a plurality of groups of training samples into a training set, a verification set and a test set according to a preset proportion; determining N preset classification models, and training the preset classification models through a training set for each preset classification model to obtain an initial model, wherein N is a positive integer; adjusting model parameters of the initial model, and determining a generalization capability evaluation value of the initial model after each adjustment of the model parameters through a verification set to obtain a group of generalization capability evaluation values of the initial model; determining an initial model corresponding to a maximum generalization capability evaluation value in a set of generalization capability evaluation values as a pending model; determining performance evaluation values of the undetermined models corresponding to each preset classification model through the test set to obtain a group of performance evaluation values of N undetermined models; and determining a pending model corresponding to the maximum performance evaluation value in the group of performance evaluation values as a target model.

Specifically, in order to train an optimal target model, a model with the maximum performance evaluation value is screened from a plurality of preset classification models, and meanwhile, the model is optimized based on the generalization capability evaluation value, for example, a plurality of groups of training samples are divided according to preset proportions, such as a 70% training set, a 15% verification set and a 15% test set, or the proportions are adjusted according to the number of samples and the complexity of the model. Model parameters such as learning rate, tree depth, neural network layer number and the like are adjusted through the verification set so as to optimize model performance, and a pending model corresponding to the maximum generalization capability evaluation value of each preset classification model is obtained. The preset classification model is such as a logistic regression model, a decision tree model, a random forest model, etc. The performance of different preset classification models is compared through cross-validation of the test set by comparing the preset classification models. Appropriate performance metrics, such as accuracy, precision, etc., are selected to evaluate the performance of the model under determination. And determining a pending model corresponding to the maximum performance evaluation value in the group of performance evaluation values as a target model. The trained object model is saved as a file for use in a production environment. Meanwhile, the target model is integrated into a monitoring system of the Internet of things equipment, so that real-time data classification and anomaly detection are realized.

Optionally, in the anomaly detection method of the internet of things device provided in the embodiment of the present application, when a newly added internet of things device exists, the method further includes: acquiring detection signals of a target sensor on newly-added internet of things equipment in an abnormal state, obtaining abnormal detection signals, and extracting a group of abnormal detection signal characteristics through the abnormal detection signals; collecting detection signals of a target sensor on newly-added internet of things equipment in a normal state, obtaining normal detection signals, and extracting a group of normal detection signal characteristics through the normal detection signals; determining a group of abnormal detection signal characteristics and abnormal detection signals as a group of newly added training samples, and determining a group of normal detection signal characteristics and normal detection signals as a group of newly added training samples; combining the newly added training samples with a plurality of groups of training samples to obtain a plurality of groups of updated training samples; and training the target model based on the updated multiple groups of training samples to obtain an updated target model.

Specifically, to update the target model in real-time, new data is collected over time, requiring periodic retraining of the target model to accommodate the change in data. For example, under the condition that newly-increased internet of things equipment exists, the abnormal detection signals and the normal detection signals of the newly-increased internet of things equipment are collected again, newly-increased training samples are determined based on a set of normal detection signal characteristics and normal detection signals, and newly-increased training samples are determined based on a set of abnormal detection signal characteristics and abnormal detection signals. And retraining the target model by adding a training sample to obtain an updated target model.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.

The embodiment of the application also provides an abnormality detection device of the internet of things device, and it is to be noted that the abnormality detection device of the internet of things device of the application embodiment can be used for executing the abnormality detection method for the internet of things device provided by the application embodiment. The following describes an abnormality detection device for an internet of things device provided in an embodiment of the present application.

Fig. 3 is a schematic diagram of an abnormality detection device of an internet of things device according to an embodiment of the present application. As shown in fig. 3, the apparatus includes:

the acquisition unit 301 is configured to acquire a detection signal of a target sensor on a target internet of things device, where the target sensor at least includes one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor;

the processing unit 302 is configured to perform noise reduction processing on the detection signal to obtain a target signal, and extract signal features from the target signal to obtain a set of signal features to be detected;

The input unit 303 is configured to input a set of signal features to be detected into a target model, to obtain a detection result for characterizing whether an abnormality exists in the target internet of things device, where the target model is obtained by training a plurality of sets of training samples, and each set of training samples includes a set of historical signal features and a set of historical detection results of the historical detection signals.

According to the anomaly detection device of the Internet of things equipment, the detection signals of the target sensor on the target Internet of things equipment are collected through the collection unit 301, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; the processing unit 302 performs noise reduction processing on the detection signal to obtain a target signal, and extracts signal characteristics from the target signal to obtain a group of signal characteristics to be detected; the input unit 303 inputs a set of signal features to be detected into the target model to obtain a detection result for representing whether the target internet of things device is abnormal, wherein the target model is obtained by training a plurality of sets of training samples, each set of training samples comprises a set of historical signal features and historical detection results of the historical detection signals, the problem that the detection efficiency of the abnormal state of the internet of things device in the related art is low is solved, the signal features to be detected are extracted through the acquired detection signals of the internet of things device, the signal features to be detected are input into the target model to obtain the detection result of whether the internet of things device is abnormal, and then the effect of improving the detection efficiency of the abnormal state of the internet of things device is achieved.

Optionally, in the abnormality detection apparatus for an internet of things device provided in the embodiment of the present application, the processing unit 302 includes: the analysis module is used for analyzing the target signal according to the time sequence to obtain a periodic waveform of the target signal; the first extraction module is used for extracting the statistical characteristics, the energy characteristics and the frequency domain characteristics of the target signal from the periodic waveform; the first determining module is configured to determine a set of signal characteristics to be detected of the target signal, where the set of signal characteristics to be detected includes at least one of: statistical features, energy features, and frequency domain features.

Optionally, in the abnormality detection apparatus for an internet of things device provided in the embodiment of the present application, the apparatus further includes: the prompt unit is used for sending prompt information when the detection result represents that the target internet of things equipment is abnormal, wherein the prompt information is used for prompting a user that the target internet of things equipment needs to be maintained; the first determining unit is used for determining the abnormal state grade of the target internet of things equipment and executing preset protection measures on the target internet of things equipment based on the abnormal state grade.

Optionally, in the abnormality detection apparatus for an internet of things device provided in the embodiment of the present application, the determining unit includes: the second extraction module is used for extracting an abnormal state evaluation value of the target internet of things device from the detection result and determining a target abnormal state evaluation range to which the abnormal state evaluation value belongs, wherein the target internet of things device is provided with a plurality of abnormal state grades in advance, and each abnormal state grade corresponds to one abnormal state evaluation range; the second determining module is configured to determine a target abnormal state level corresponding to the target abnormal state evaluation range, and execute a preset protection measure corresponding to the target abnormal state level on the target internet of things device, where the preset protection measure at least includes one of the following: pausing the function, shutdown, power outage and networking diagnostics of the target internet of things device.

Optionally, in the abnormality detection apparatus for an internet of things device provided in the embodiment of the present application, the acquisition unit 301 includes: the third determining module is configured to determine a device type to which the target internet of things device belongs, and determine, through the device type, a target sensor required for acquiring a detection signal of the target internet of things device from a preset corresponding relation set, where the preset corresponding relation set includes a plurality of corresponding relations, and each corresponding relation includes an internet of things device of one device type and at least one sensor for acquiring the detection signal of the internet of things device.

Optionally, in the anomaly detection device for an internet of things device provided in the embodiment of the present application, the target model is obtained by: acquiring a plurality of history detection signals, and extracting a group of history signal characteristics from each history detection signal; determining a history detection result of each history detection signal, and taking a group of history signal characteristics and the history detection result of each history detection signal as a group of training samples to obtain a plurality of groups of training samples; training a preset classification model based on a plurality of groups of training samples to obtain a target model, wherein the preset classification model at least comprises one of the following: logistic regression models, decision tree models, and random forest models.

Optionally, in the anomaly detection device of the internet of things device provided in the embodiment of the present application, training a preset classification model based on multiple sets of training samples, obtaining the target model includes: dividing a plurality of groups of training samples into a training set, a verification set and a test set according to a preset proportion; determining N preset classification models, and training the preset classification models through a training set for each preset classification model to obtain an initial model, wherein N is a positive integer; adjusting model parameters of the initial model, and determining a generalization capability evaluation value of the initial model after each adjustment of the model parameters through a verification set to obtain a group of generalization capability evaluation values of the initial model; determining an initial model corresponding to a maximum generalization capability evaluation value in a set of generalization capability evaluation values as a pending model; determining performance evaluation values of the undetermined models corresponding to each preset classification model through the test set to obtain a group of performance evaluation values of N undetermined models; and determining a pending model corresponding to the maximum performance evaluation value in the group of performance evaluation values as a target model.

Optionally, in the abnormality detection apparatus for an internet of things device provided in the embodiment of the present application, the apparatus further includes: the first extraction unit is used for acquiring detection signals of a target sensor on the newly-added internet of things equipment in an abnormal state, obtaining abnormal detection signals and extracting a group of abnormal detection signal characteristics through the abnormal detection signals; the second extraction unit is used for acquiring detection signals of the target sensor on the newly-added internet of things equipment in a normal state, obtaining normal detection signals and extracting a group of normal detection signal characteristics through the normal detection signals; the second determining unit is used for determining a group of abnormal detection signal characteristics and abnormal detection signals as a group of newly added training samples, and determining a group of normal detection signal characteristics and normal detection signals as a group of newly added training samples; the combining unit is used for combining the newly added training samples with a plurality of groups of training samples to obtain a plurality of groups of updated training samples; and the training unit is used for training the target model based on the updated multiple groups of training samples to obtain an updated target model.

The abnormality detection device of the internet of things device includes a processor and a memory, the above-mentioned acquisition unit 301, processing unit 302, input unit 303, and the like are stored as program units in the memory, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel can be provided with one or more than one, and the abnormal state detection efficiency of the Internet of things equipment is improved by adjusting kernel parameters.

The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.

The embodiment of the invention provides a computer readable storage medium, on which a program is stored, which when executed by a processor, implements an anomaly detection method for an internet of things device.

The embodiment of the invention provides a processor, which is used for running a program, wherein the program runs to execute an abnormality detection method of Internet of things equipment.

Fig. 4 is a schematic diagram of an electronic device provided according to an embodiment of the present application. As shown in fig. 4, the electronic device 401 includes a processor, a memory, and a program stored on the memory and executable on the processor, and the processor implements the following steps when executing the program: collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment has abnormality, wherein the target model is trained by a plurality of groups of training samples, and each group of training samples comprises a group of historical signal characteristics and a historical detection result of a historical detection signal. The device herein may be a server, PC, PAD, cell phone, etc.

The present application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor; noise reduction processing is carried out on the detection signals to obtain target signals, and signal characteristics are extracted from the target signals to obtain a group of signal characteristics to be detected; inputting a group of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target Internet of things equipment has abnormality, wherein the target model is trained by a plurality of groups of training samples, and each group of training samples comprises a group of historical signal characteristics and a historical detection result of a historical detection signal.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.

Claims (10)

1. The abnormality detection method for the Internet of things equipment is characterized by comprising the following steps of:

collecting detection signals of a target sensor on target internet of things equipment, wherein the target sensor at least comprises one of the following: a power sensor, a voltage sensor, a current sensor and a temperature sensor;

carrying out noise reduction treatment on the detection signals to obtain target signals, and extracting signal characteristics from the target signals to obtain a group of signal characteristics to be detected;

inputting the set of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target internet of things equipment has abnormality, wherein the target model is obtained by training a plurality of sets of training samples, and each set of training samples comprises a set of historical signal characteristics and a set of historical detection results of historical detection signals.

2. The method of claim 1, wherein extracting signal features from the target signal to obtain a set of signal features to be detected comprises:

analyzing the target signal according to a time sequence to obtain a periodic waveform of the target signal;

extracting statistical features, energy features and frequency domain features of the target signal from the periodic waveform;

Determining a set of signal characteristics to be detected of the target signal, wherein the set of signal characteristics to be detected comprises at least one of: the statistical features, the energy features, and the frequency domain features.

3. The method of claim 1, wherein after obtaining a detection result for characterizing whether the target internet of things device is abnormal, the method further comprises:

under the condition that the detection result represents that the target Internet of things equipment is abnormal, prompt information is sent out, wherein the prompt information is used for prompting a user that the target Internet of things equipment needs maintenance;

and determining the abnormal state grade of the target Internet of things equipment, and executing preset protection measures on the target Internet of things equipment based on the abnormal state grade.

4. The method of claim 3, wherein determining an abnormal state level of the target internet of things device and performing a preset protection measure on the target internet of things device based on the abnormal state level comprises:

extracting an abnormal state evaluation value of the target internet of things device from the detection result, and determining a target abnormal state evaluation range to which the abnormal state evaluation value belongs, wherein the target internet of things device is provided with a plurality of abnormal state grades in advance, and each abnormal state grade corresponds to one abnormal state evaluation range;

Determining a target abnormal state grade corresponding to the target abnormal state evaluation range, and executing a preset protection measure corresponding to the target abnormal state grade on the target internet of things equipment, wherein the preset protection measure at least comprises one of the following steps: and pausing the functions, shutdown, power outage and networking diagnosis of the target Internet of things equipment.

5. The method of claim 1, wherein acquiring detection signals of a target sensor on a target internet of things device comprises:

determining the equipment type of the target Internet of things equipment, and determining a target sensor required for acquiring detection signals of the target Internet of things equipment from a preset corresponding relation set through the equipment type, wherein the preset corresponding relation set comprises a plurality of corresponding relations, and each corresponding relation comprises the Internet of things equipment of one equipment type and at least one sensor for acquiring the detection signals of the Internet of things equipment.

6. The method of claim 1, wherein the target model is derived by:

acquiring a plurality of history detection signals, and extracting a group of history signal characteristics from each history detection signal;

Determining a history detection result of each history detection signal, and taking a group of history signal characteristics and the history detection result of each history detection signal as a group of training samples to obtain a plurality of groups of training samples;

training a preset classification model based on the plurality of groups of training samples to obtain the target model, wherein the preset classification model at least comprises one of the following: logistic regression models, decision tree models, and random forest models.

7. The method of claim 6, wherein training a preset classification model based on the plurality of sets of training samples to obtain the target model comprises:

dividing the multiple groups of training samples into a training set, a verification set and a test set according to a preset proportion;

determining N preset classification models, and training the preset classification models through the training set for each preset classification model to obtain an initial model, wherein N is a positive integer;

adjusting model parameters of the initial model, and determining a generalization capability evaluation value of the initial model after each adjustment of the model parameters through the verification set to obtain a set of generalization capability evaluation values of the initial model;

determining an initial model corresponding to the maximum generalization capability evaluation value in the set of generalization capability evaluation values as a pending model;

Determining performance evaluation values of the undetermined models corresponding to each preset classification model through the test set to obtain a group of performance evaluation values of N undetermined models;

and determining a pending model corresponding to the maximum performance evaluation value in the group of performance evaluation values as the target model.

8. The method of claim 1, wherein in the event that there is a newly added internet of things device, the method further comprises:

acquiring detection signals of a target sensor on newly-added internet of things equipment in an abnormal state, obtaining abnormal detection signals, and extracting a group of abnormal detection signal characteristics through the abnormal detection signals;

acquiring detection signals of a target sensor on newly-added internet of things equipment in a normal state to obtain normal detection signals, and extracting a group of normal detection signal characteristics through the normal detection signals;

determining the set of abnormal detection signal characteristics and the abnormal detection signal as a set of newly added training samples, and determining the set of normal detection signal characteristics and the normal detection signal as a set of newly added training samples;

combining the newly added training samples with the multiple groups of training samples to obtain updated multiple groups of training samples;

And training the target model based on the updated multiple groups of training samples to obtain an updated target model.

9. An anomaly detection device of an internet of things device, comprising:

the system comprises an acquisition unit, a detection unit and a control unit, wherein the acquisition unit is used for acquiring detection signals of a target sensor on target internet of things equipment, and the target sensor at least comprises one of the following components: a power sensor, a voltage sensor, a current sensor and a temperature sensor;

the processing unit is used for carrying out noise reduction processing on the detection signals to obtain target signals, extracting signal characteristics from the target signals and obtaining a group of signal characteristics to be detected;

the input unit is used for inputting the set of signal characteristics to be detected into a target model to obtain a detection result used for representing whether the target internet of things equipment has an abnormality, wherein the target model is trained by a plurality of sets of training samples, and each set of training samples comprises a set of historical signal characteristics and a set of historical detection result of the historical detection signal.

10. An electronic device comprising one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the anomaly detection method of the internet of things device of any one of claims 1 to 8.