Intelligent voltage protection system and method based on Internet of Things

Abstract

The application provides an intelligent voltage protection system and method based on the Internet of Things. The application performs device action recognition according to behavior correlation data to obtain a device external action recognition result of a target device, performs device action recognition according to internal voltage state data to obtain an internal action recognition result of the target device, performs behavior consistency verification according to the device external action recognition result and the internal action recognition result to obtain an external behavior consistency degree, performs dynamic behavior prediction on the target device according to image sensing data and the external behavior consistency degree to determine a device behavior prediction result, dynamically adjusts a device voltage protection threshold according to the behavior prediction result, and performs intelligent voltage protection on the target device by the device voltage protection threshold. The target device behavior can be predicted through the image sensing data of the Internet of Things, and the voltage protection threshold is dynamically adjusted according to the behavior prediction result, so that voltage false alarms caused by frequent start and stop are avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of voltage protection, and more particularly, to an intelligent voltage protection system and method based on Internet of Things. BACKGROUND

[0002] Voltage protection is a relay protection mode that takes the measured voltage at the installation site of the device as the action quantity. The relay protection formed by the excessively low or high voltage is respectively called low voltage protection and overvoltage protection. The operating voltage of the power system is an important indicator of power quality. When it deviates from the rated voltage by more than a certain limit, it may damage the stable operation of the system, affect the quantity and quality of factory products, and also cause damage to the electrical equipment. With the progress of voltage protection technology, the relay protection calculation intelligent analysis and early warning system based on Internet of Things and big data emerges as the times require, providing new technical support for the stable operation of the power system.

[0003] In the prior art, the electrical parameter detection of the target device is mainly performed through Internet of Things, and then voltage protection is performed through fixed threshold judgment to avoid the operating voltage of the target device deviating from the rated voltage by more than a certain limit. However, during the period of industrial power consumption concentration, due to the frequent start-stop of large equipment such as electric motors, air compressors, and welding equipment, short-term fluctuations or even impact changes in current and voltage may be caused. Such fluctuations are transient and predictable, and the fixed voltage protection threshold is easy to misjudge these short-term changes as voltage abnormalities, triggering voltage false alarms. Therefore, how to solve the voltage false alarm caused by the frequent start-stop of the target device has become a problem to be solved. SUMMARY

[0004] The present application provides an intelligent voltage protection system and method based on Internet of Things, which can predict the behavior of the target device through Internet of Things image sensing data, and dynamically adjust the voltage protection threshold according to the behavior prediction result, thereby avoiding voltage false alarms caused by frequent start-stop.

[0005] In a first aspect, the present application provides an intelligent voltage protection method based on Internet of Things. The method can be executed by a network device, or a chip configured in the network device. The present application does not limit this.

[0006] Specifically, the method comprises:

[0007] Collecting external behavior correlation data of the target device through Internet of Things, performing device action recognition according to the external behavior correlation data, and obtaining a device external action recognition result of the target device;

[0008] Collecting internal voltage state data of the target device, performing device action recognition according to the internal voltage state data, and obtaining an internal action recognition result of the target device;

[0009] perform behavior consistency verification according to the device external action recognition result and the internal action recognition result, to obtain an external behavior consistency degree;

[0010] When the external behavior consistency degree is higher than a preset trust threshold, image sensor data of the target device is acquired, behavior dynamic prediction is performed on the target device according to the image sensor data and the external behavior consistency degree, and a device behavior prediction result is determined;

[0011] According to the behavior prediction result, a device voltage protection threshold is dynamically adjusted, and intelligent voltage protection is performed on the target device by the device voltage protection threshold.

[0012] In combination with the first aspect, in some implementations of the first aspect, the internal voltage state data is collected by a voltage sensor installed in a power supply line circuit of the target device.

[0013] In combination with the first aspect, in some implementations of the first aspect, the device action recognition according to the internal voltage state data specifically includes:

[0014] According to the internal voltage state data, a time sequence sample set of the voltage signal is constructed by using a sliding time window.

[0015] Voltage feature extraction is performed on the voltage data in each time window, and a voltage feature vector is constructed according to a plurality of voltage features corresponding to each time window, to obtain a feature vector set.

[0016] Each voltage feature vector in the feature vector set is classified by a trained neural network, to obtain an internal action state label corresponding to each time window, and to obtain the internal action recognition result.

[0017] In combination with the first aspect, in some implementations of the first aspect, the behavior consistency verification according to the device external action recognition result and the internal action recognition result specifically includes:

[0018] According to the device external action recognition result label corresponding to each time window, an encoding value mapping is performed, to obtain a corresponding external action recognition encoding sequence.

[0019] According to the device internal action recognition result label corresponding to each time window, an encoding value mapping is performed, to obtain a corresponding internal action recognition encoding sequence.

[0020] Based on the external action recognition encoding sequence and the internal action recognition encoding sequence, consistency verification is performed, to determine the external behavior consistency degree.

[0021] With reference to the first aspect, in some implementations of the first aspect, the behavior dynamic prediction of the target device according to the image sensing data and the external behavior consistency degree comprises:

[0022] After the data preprocessing of the image sensing data, the time-series image features are extracted through a convolutional neural network;

[0023] The behavior prediction depth of the target device is determined according to the external behavior consistency, the time-series image features are modeled through moving average autoregression according to the behavior prediction depth, and the device behavior prediction result of the target device is determined through the moving average autoregression model.

[0024] With reference to the first aspect, in some implementations of the first aspect, the external behavior correlation data comprises external image data, external sound data and external vibration data of the target device.

[0025] With reference to the first aspect, in some implementations of the first aspect, the device action recognition according to the external behavior correlation data comprises:

[0026] The external image data, the external sound data and the external vibration data in the behavior correlation data are acquired;

[0027] The image action recognition is performed on the external image data to determine a first action recognition result;

[0028] The device sound recognition is performed on the external sound data to determine a second action recognition result;

[0029] The vibration anomaly recognition is performed on the external vibration data to determine a third action recognition result;

[0030] The first, second and third action recognition results are fused to determine the device external action recognition result of the target device.

[0031] The second aspect, the application provides a kind of based on the intelligent voltage protection system of Internet of Things, it includes voltage protection unit, the voltage protection unit includes:

[0032] Device action recognition module, for collecting the external behavior correlation data of target device by Internet of Things, according to the external behavior correlation data, device action recognition is carried out, and the device external action recognition result of the target device is obtained;

[0033] The device action recognition module is also used to collect the internal voltage state data of target device, according to the internal voltage state data, device action recognition is carried out, and the internal action recognition result of the target device is obtained;

[0034] a decision module configured to perform behavior consistency verification according to the external action recognition result and the internal action recognition result, and obtain an external behavior consistency degree;

[0035] a device behavior prediction module configured to, when the external behavior consistency degree is higher than a preset trust threshold, acquire image sensing data of the target device, perform dynamic behavior prediction on the target device according to the image sensing data and the external behavior consistency degree, and determine a device behavior prediction result;

[0036] a voltage protection adjustment module configured to dynamically adjust a device voltage protection threshold according to the behavior prediction result, and perform intelligent voltage protection on the target device by the device voltage protection threshold.

[0037] In a third aspect, the present application provides a computer terminal device, which comprises a memory and a processor, the memory stores a code, and the processor is configured to acquire the code and execute the above-mentioned smart voltage protection method based on Internet of Things.

[0038] In a fourth aspect, the present application provides a computer readable storage medium, which stores at least one computer program, the computer program is loaded and executed by a processor to realize the operation performed by the above-mentioned smart voltage protection method based on Internet of Things.

[0039] The technical scheme provided by the embodiments of the present application has the following beneficial effects:

[0040] In the smart voltage protection system and method based on Internet of Things provided by the present application, external behavior correlation data of a target device is collected through Internet of Things, device action recognition is performed according to the behavior correlation data, and an external action recognition result of the target device is obtained; internal voltage state data of the target device is collected, device action recognition is performed according to the internal voltage state data, and an internal action recognition result of the target device is obtained; behavior consistency verification is performed according to the external action recognition result and the internal action recognition result, and an external behavior consistency degree is obtained; when the external behavior consistency degree is higher than a preset trust threshold, image sensing data of the target device is acquired, dynamic behavior prediction is performed on the target device according to the image sensing data and the external behavior consistency degree, and a device behavior prediction result is determined; and dynamic adjustment is performed on a device voltage protection threshold according to the behavior prediction result, and intelligent voltage protection is performed on the target device by the device voltage protection threshold.

[0041] It can be seen that the application can collect external behavior correlation data through Internet of Things sensors, accurately identify whether the device is in a running, starting, stopping or abnormal state from a multi-modal perspective, avoid the risk of misjudging the device state by a single voltage signal, improve the understanding ability of the system to the real device behavior, perform consistency test on the external behavior recognition result and the internal voltage state recognition result, and calculate the external behavior consistency degree, so as to ensure that only when the external observation result is consistent with the actual voltage state, the in-depth prediction is performed, fundamentally reduce the false prediction and protection failure caused by the perception deviation, and adaptively adjust the device behavior prediction according to the external behavior consistency degree, improve the prediction flexibility and fault tolerance, reduce the misjudgment risk, compared with the traditional method of relying only on voltage fluctuation for judgment, the application can capture the pre-sequence characteristics of device running change in advance through image sensing data, such as motor rotation initial motion, indicator light lighting, operation panel change, etc., so as to judge the device state in advance, gain reaction time for dynamic threshold adjustment, avoid false alarm caused by threshold lag, so as to reasonably improve or reduce the voltage alarm threshold according to the prediction result in the device frequent start-stop scene, improve the precision and robustness of the voltage protection system.

[0042] In summary, the application can perform target device behavior prediction through Internet of Things image sensing data, and dynamically adjust the voltage protection threshold according to the behavior prediction result, so as to avoid voltage false alarm caused by frequent start-stop. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is an exemplary flowchart of an intelligent voltage protection method based on Internet of Things according to some embodiments of the application;

[0044] Figure 2 is a structural schematic diagram of a voltage protection unit according to some embodiments of the application;

[0045] Figure 3 is a structural schematic diagram of a computer terminal device for implementing an intelligent voltage protection method based on Internet of Things according to some embodiments of the application. DETAILED DESCRIPTION

[0046] The application collects external behavior correlation data of a target device through the Internet of Things, performs device action recognition according to the behavior correlation data, obtains a device external action recognition result of the target device, collects internal voltage state data of the target device, performs device action recognition according to the internal voltage state data, obtains an internal action recognition result of the target device, performs behavior consistency verification according to the device external action recognition result and the internal action recognition result, obtains an external behavior consistency degree, when the external behavior consistency degree is higher than a preset trust threshold, acquires image sensing data of the target device, performs behavior dynamic prediction on the target device according to the image sensing data and the external behavior consistency degree, and determines a device behavior prediction result, dynamically adjusts a device voltage protection threshold according to the behavior prediction result, and performs intelligent voltage protection on the target device by the device voltage protection threshold, so that target device behavior prediction can be performed through Internet of Things image sensing data, and the voltage protection threshold is dynamically adjusted according to the behavior prediction result, thereby avoiding frequent start-stop caused voltage false alarm.

[0047] In order to better understand the above technical solutions, the above technical solutions will be described in detail below in combination with the drawings of the specification and specific embodiments. Reference Figure 1 The figure is an example flow chart of an intelligent voltage protection method based on the Internet of Things according to some embodiments of the application, which mainly includes the following steps:

[0048] In step S101, external behavior correlation data of a target device is collected through the Internet of Things, device action recognition is performed according to the external behavior correlation data, and a device external action recognition result of the target device is obtained.

[0049] It should be noted that the external behavior correlation data is obtained in real time by the Internet of Things sensor network installed around the target device, which is used for non-intrusive identification and state judgment of the running behavior of the target device. The external behavior correlation data at least includes external image data, external sound data and external vibration data of the target device, wherein the external image data includes image sequences or video streams of the target device in running or standby state, which is used to analyze whether the device has visible action change (such as rotation, movement, opening, closing, etc.); the external sound data is the running noise generated when the target device works, for example, high-frequency noise when the motor starts, continuous running sound, etc., which is used to reflect the dynamic state of the device, and the external vibration data is the mechanical vibration generated in the running process of the device obtained by three-axis acceleration sensor, MEMS sensor, etc., which is used to detect whether there is device start, load fluctuation or abnormal vibration.

[0050] Preferably, in some embodiments, the device action recognition according to the external behavior correlation data to obtain the device external action recognition result of the target device specifically includes:

[0051] acquiring external image data, external sound data, and external vibration data from the behavior-related data;

[0052] performing image action recognition on the external image data to determine a first action recognition result;

[0053] performing device sound recognition on the external sound data to determine a second action recognition result;

[0054] performing vibration abnormality recognition on the external vibration data to determine a third action recognition result;

[0055] A fusion decision is performed based on the first, second, and third action recognition results to determine an external device action recognition result of the target device.

[0056] In specific implementation, the action recognition of external image data can be achieved by introducing the YOLOV8 target detection model, performing continuous frame difference analysis and feature extraction on the external image data, identifying the motion state of the device in the image, and combining optical flow method, edge detection, regional motion vector and other algorithms to determine whether the target device is in a motion state; the recognition result outputs the corresponding action state label, such as "run", "stop" or "unknown", as the first action recognition result.

[0057] During the action recognition process of external sound data, the collected external sound signal is subjected to Fourier transform or Mel-frequency cepstral coefficient extraction to generate spectrum graph data during the operation of the device. The spectrum data is input into the trained long-short-term memory network model for dynamic time series modeling and cluster analysis, and the current action state label of the motor or device, such as "standby", "starting", "stable operation", etc., is output as the second action recognition result.

[0058] Vibration recognition: By performing wavelet analysis or bandpass filtering on the external vibration data, and extracting time domain and frequency domain features of key frequency bands, including characteristic values ​​such as acceleration mean, frequency peak, and short-time energy, the data are input into a one-dimensional convolutional neural network or other clustering model to complete vibration pattern recognition and output the corresponding vibration action label as the third action recognition result.

[0059] To improve recognition accuracy, the first, second and third action recognition results will be aggregated into a fusion decision module for unified decision making.

[0060] Preferably, the fusion module adopts a confidence-based weighted decision strategy, in which: each action recognition model will calibrate its confidence output during the training process; based on historical verification experiments, dataset quality and model stability, the system assigns fixed confidence weights to image recognition, sound recognition and vibration recognition, for example:

[0061] Image recognition: weight 0.4; sound recognition: weight 0.35; vibration recognition: weight 0.25;

[0062] The final recognition result is fused and calculated according to the confidence weighted score to determine the current target device external behavior recognition state.

[0063] For example, if the image model recognizes the device as "running" with a confidence of 0.92, the sound model as "running" with a confidence of 0.87, and the vibration model as "stop" with a confidence of 0.78, the fusion decision is "running" with a confidence weight calculated as follows: fusion score = (0.4 x 0.92) + (0.35 x 0.87) + (0.25 x 0.22) = 0.748, and the system can set a fusion threshold (such as 0.65), and if it exceeds, it is determined as the current effective recognition state.

[0064] In step S102, internal voltage state data of the target device is collected, and device action recognition is performed according to the internal voltage state data to obtain an internal action recognition result of the target device.

[0065] In a preferred embodiment of the present application, the internal voltage state data refers to voltage change data reflecting the electrical working state of the device, which is monitored and recorded in real time by a voltage collection module installed in the power input end, control panel or electric control cabinet of the target device, and is used to identify the internal running behavior of the device, including whether it is starting, running, stopping or having an abnormal state, etc.

[0066] Optionally, in some embodiments, the internal voltage state data is collected in real time by a voltage sensor (which can be selected as a voltage transformer, a Hall voltage sensor, a Rogowski coil, a voltage dividing circuit, etc.) installed in the power input circuit of the target device, and the sampling accuracy is not less than ±1V; in other embodiments, the voltage signal can also be digitized by an analog-to-digital converter module, and processed by a microprocessor, an edge computing node or an embedded terminal. The sampling frequency can be set according to the response characteristics of the target device, and is preferably 1-10 kHz. The present application does not limit the type, communication mode (such as Modbus, CAN, RS485, Wi-Fi) etc. of the specific voltage collection hardware, and the specific implementation can be adapted to different industrial scenes and Internet of Things platform architectures.

[0067] Preferably, in some embodiments, the internal action recognition result of the target device obtained according to the internal voltage state data specifically includes:

[0068] According to the internal voltage state data, a time sequence sample set of the voltage signal is constructed by using a sliding time window;

[0069] extract voltage features from the voltage data in each time window, and construct a voltage feature vector according to the plurality of voltage features corresponding to each time window, to obtain a feature vector set;

[0070] Each voltage feature vector in the feature vector set is classified by the trained neural network to obtain an internal action state label corresponding to each time window, and an internal action recognition result is obtained.

[0071] In a specific implementation, based on the continuously collected internal voltage state data, a fixed length sliding time window (such as 1s, 2s or adaptive) is set to construct a time series sample set of the voltage signal. Each time window forms a local voltage data segment containing a plurality of voltage sampling points, and then multi-dimensional feature extraction is performed on the voltage data in each time window, including but not limited to: time domain features: maximum value, minimum value, mean value, standard deviation, slope change rate; frequency domain features: FFT spectrum peak value, frequency center, bandwidth; mutation features: zero-crossing number, fluctuation amplitude, peak number; energy features: short-time energy, waveform entropy, power estimation; the features extracted in each window constitute a voltage feature vector, and a plurality of time windows can form a feature vector set. The above feature vectors are input into a pre-trained neural network model (such as a convolutional neural network CNN), and the model identifies the state of the current time window according to the classification boundary and distribution law learned from history, and outputs an internal action state label corresponding to the current time window.

[0072] The internal action state label can include "start", "run", "stop" and "abnormal", etc. In a specific application example of the present application, the rated voltage of an industrial motor device is 380V: when the motor starts, the voltage is collected from 380V to 330V, and recovers within 0.8 seconds. This mode is identified as a typical start behavior by the system; after starting, the voltage remains in the fluctuation range of 376-384V and lasts for 5 minutes, then the system judges it as "running"; if the voltage is 0 for 10 seconds, it is determined that the device has "stopped"; if the voltage waveform frequently mutates and more than 10 peak disturbances occur within 20ms, the "abnormal" state is output.

[0073] In step S103, behavior consistency verification is performed according to the device external action recognition result and the internal action recognition result, to obtain an external behavior consistency degree.

[0074] It should be noted that the external behavior consistency degree in the present application refers to the consistency between the recognition results of the external perception data (such as images, sounds, vibrations) of the target device and the recognition results of the internal running state (such as voltage change), so as to measure the matching degree between the external performance of the device and the actual action. When the external behavior consistency degree is higher than a preset threshold, it indicates that the external perception data of the device is highly reliable, and the dynamic behavior prediction of the target device can be performed based on the external perception data, and the dynamic adjustment of the voltage protection threshold is performed according to the behavior prediction result.

[0075] Preferably, in some embodiments, the behavior consistency verification is performed according to the device external action recognition result and the internal action recognition result, and the external behavior consistency degree is obtained, which specifically includes:

[0076] According to the encoding value mapping of the device external action recognition result label corresponding to each time window, the corresponding external action recognition encoding sequence is obtained;

[0077] According to the encoding value mapping of the device internal action recognition result label corresponding to each time window, the corresponding internal action recognition encoding sequence is obtained;

[0078] The consistency verification is performed based on the external action recognition encoding sequence and the internal action recognition encoding sequence, and the external behavior consistency degree is determined.

[0079] In specific implementation, in order to make the recognition result applicable to the correlation coefficient analysis, the "external action state label" and the "internal action state label" need to be uniformly encoded into numerical variables, for example, the numerical encoding of the "start" label is 1, the numerical encoding of the "run" label is 2, the numerical encoding of the "stop" label is 0, and the numerical encoding of the "abnormal" label is -1.

[0080] In the present application, the sliding window length is taken as an example of 1S, the external action recognition encoding sequence and the internal action recognition encoding sequence of the target device are obtained at the same time, and the Pearson correlation coefficient is determined based on the external action recognition encoding sequence and the internal action recognition encoding sequence as the external behavior consistency degree.

[0081] In a specific application scenario of the present application, in the time sequence in the past 10 seconds, the external action recognition encoding sequence is X=[1, 1, 2, 2, 2, 2, 2, 0, 0, 0] (start→run→stop), and the internal action recognition encoding sequence is Y=[1, 2, 2, 2, 2, 2, 1, 0, 0, 0] (start slightly faster, run short-term fluctuation). The Pearson coefficient is calculated as: ρ=0.97. Therefore, it is judged that the external behavior and the internal state are highly consistent, and it is considered that the recognition result is reliable, which can be used as the basis for subsequent behavior prediction and voltage protection adjustment.

[0082] In step S104, when the external behavior consistency is higher than a preset trust threshold, image sensing data of the target device is acquired, and the target device is dynamically predicted in behavior according to the image sensing data and the external behavior consistency, and a device behavior prediction result is determined.

[0083] Optionally, in some embodiments, when the external behavior consistency is lower than the preset trust threshold, since there is a large difference between the external behavior recognition result and the internal voltage state, it is indicated that the current data can be unreliable or abnormal, the system can suspend the dynamic prediction of behavior based on the external behavior data, avoid misleading the subsequent threshold adjustment, in the case that the behavior prediction is unavailable or untrusted, the system returns to the preset static voltage protection threshold range, ensures that the basic function of the protection mechanism is not affected, and prevents misoperation or false alarm.

[0084] Preferably, in some embodiments, the image sensing data of the target device can be collected by an industrial camera, and in some other embodiments, it can also be collected by other devices or equipment capable of realizing image collection of the target device, and the present application does not limit this.

[0085] Preferably, in some embodiments, the dynamic prediction of behavior of the target device according to the image sensing data and the external behavior consistency to determine the device behavior prediction result specifically includes:

[0086] After data preprocessing of the image sensing data, time series image features are extracted through a convolutional neural network;

[0087] The behavior prediction depth of the target device is determined according to the external behavior consistency, a moving average autoregressive model is established for the time series image features according to the behavior prediction depth, and the device behavior prediction result of the target device is determined through the moving average autoregressive model.

[0088] In specific implementation, the collected image sensing data is standardized and preprocessed to improve the accuracy and stability of subsequent feature extraction and modeling. The preprocessing steps include but are not limited to: image graying, denoising; size scaling and image enhancement, which can use histogram equalization and edge enhancement; time alignment of image frames.

[0089] It should be noted that the present application introduces external behavior consistency as a regulating factor in the device behavior prediction process, which can effectively improve the accuracy and reliability of the prediction. When the external behavior consistency is high, it indicates that the external perception data and the internal running state have a high degree of matching, at this time the prediction depth can be increased, more historical feature data is used for trend modeling, thereby improving the accuracy of behavior prediction; when the consistency is low, the system can automatically shorten the prediction window or suspend the prediction, to avoid misjudgment caused by unreliable data. In addition, the mechanism can also dynamically adjust the voltage protection strategy according to the consistency, realize the fine and adaptive control of intelligent protection, reduce the consumption of computing resources, enhance the perception and response ability of the system to potential abnormal behavior, and significantly improve the stability and intelligent level of the entire voltage protection system.

[0090] Further, the preprocessed image sequence is input into a convolutional neural network model to extract the spatial-temporal feature representation corresponding to each frame or each several frames of images. The convolutional neural network model can be a trained lightweight network (such as MobileNet, ResNet-18, EfficientNet), and the extracted feature vectors are arranged in time sequence to form a feature sequence, and then the behavior prediction depth of the target device is adjusted by the external behavior consistency, specifically, according to the range interval of the external behavior consistency, a mapping table is dynamically mapped according to the preset mapping table to determine the number of image frames as the corresponding behavior prediction depth, the behavior prediction depth is the length of the moving average modeling window, and then the image feature sequence in the prediction depth range is selected as the input feature for moving average autoregressive modeling, the next period image feature vector predicted is matched with the historical label library for feature clustering (preferably using K-Means clustering algorithm), and the most possible action state corresponding to the current feature in the label space is identified. For example: cluster center A→“running”, cluster center B→“about to start”, cluster center C→“abnormal vibration”, and finally the device behavior prediction result of the target device is output, which can also be accompanied by its corresponding behavior label and prediction confidence score, for guiding the dynamic adjustment of the protection threshold of the voltage protection system.

[0091] A specific embodiment of the present application for determining the device behavior prediction result of the target device by using a moving average autoregressive model is given as follows:

[0092] The system collects image data of the target device in a period of time from the image sensor in real time, and after image preprocessing, inputs the image data into a trained lightweight convolutional neural network (such as ResNet-18) to extract the spatial-temporal feature representation of each frame of image or image sequence, and generate a continuous image feature vector sequence.

[0093] In order to further reduce the dimension and enhance the modeling feasibility, principal component analysis (PCA) can be performed on the extracted image feature sequence, and the first 1-2 principal components are retained as the image behavior index sequence to form a time series.

[0094] Further, the depth of behavior prediction is dynamically adjusted in combination with the external behavior consistency degree. In the embodiment, the following mapping rule of external behavior consistency degree and prediction depth is set: when the external behavior consistency degree p is greater than or equal to 0.9, the system considers that the current external and internal recognition are highly consistent, and the data reliability is high, at this time, the behavior prediction depth is set to 60, that is, the image features in the past 60 minutes are used for behavior modeling and prediction; when the external behavior consistency degree p is between 0.7 and 0.9, the data consistency is good, and the behavior prediction depth is set to 30; when the external behavior consistency degree p is between 0.5 and 0.7, the data consistency is general, and the prediction depth is reduced to 15; when the external behavior consistency degree p is less than 0.5, it indicates that there is a large difference between the external and internal recognition results, and the data reliability is low, at this time, long-term prediction is not recommended, and short-cycle prediction (for example, the window length is 5) can be selected, or the prediction stage is skipped.

[0095] According to the external behavior consistency degree calculated at the current time, the corresponding prediction window length is selected, that is, the corresponding number of data points in the image behavior index sequence are intercepted to form a prediction sequence.

[0096] Subsequently, the moving average autoregressive sequence is detected for stationarity (such as ADF test), and if it is not stationary, first-order difference or exponential smoothing processing is performed. Then, the autocorrelation coefficient graph (ACF) and the partial autocorrelation coefficient graph (PACF) are drawn, and the order (p, q) of the ARMA model is determined according to the significant decay points in the graph.

[0097] For example: if the ACF is truncated after 3 orders, and the PACF gradually decays after 2 orders, the order of the ARMA model can be determined as (3, 2); parameter estimation can be performed by least squares method or maximum likelihood estimation, and the confidence level is 0.05.

[0098] The ARMA (p, q) model is used to predict the image behavior index value at the next time. If the predicted value is significantly higher than the historical average value, it may indicate that the device is about to start or enter a high-load running state; if the predicted value decreases, it may indicate that the device is preparing to stop or the operation tends to be stable.

[0099] Finally, the corresponding behavior label (such as “about to start”, “continuous operation”, “preparing to stop” or “abnormal”) is matched according to the prediction result, and is used as the basis for subsequent dynamic voltage protection threshold adjustment.

[0100] It should be noted that compared with the traditional method of judging only by voltage fluctuation, the image sensing data can not only better reflect the actual running start-stop condition of the device, but also can capture the pre-sequence characteristics of the device running change in advance, such as motor rotation initial motion, indicator light lighting, operation panel change, etc., so as to judge the device state in advance, gain reaction time for dynamic threshold adjustment, avoid false alarm caused by threshold lag, and thus avoid false alarm caused by voltage short-time drop in the device frequent start-stop scene, such as reasonably increasing or reducing the voltage alarm threshold according to the prediction result (for example, temporarily relaxing the threshold range when predicting that the device will start soon), and vice versa, when predicting that the device will stop or appear abnormal fluctuation, tightening the threshold range in time to improve the protection sensitivity, effectively improving the accuracy and robustness of the voltage protection system.

[0101] In step S105, the device voltage protection threshold is dynamically adjusted according to the behavior prediction result, and the target device is intelligently protected by the device voltage protection threshold.

[0102] In a preferred embodiment of the present application, the device behavior prediction result can be used to perceive the working state of the target device in advance, such as that the device will start, the load will rise or there is a high possibility of shutdown, etc. Based on the prediction result, the voltage protection threshold of the device is dynamically adjusted to adapt to the upcoming voltage fluctuation situation, so as to improve the sensitivity and fault tolerance of the protection system.

[0103] Specifically, it can include the following steps: behavior prediction result acquisition: based on image sensing data, external behavior consistency degree and behavior modeling algorithm (such as moving average autoregressive model), the device behavior prediction result of the target device in a prediction period in the future is obtained, and the prediction result can include labels such as "about to start", "continuous running", "planned shutdown" or "possible abnormality". The threshold adjustment strategy matching will call the corresponding voltage protection strategy and dynamically adjust the voltage protection threshold according to different prediction results. For example: if the prediction result is "about to start": since the motor start is usually accompanied by voltage instantaneous drop, the system will temporarily lower the lower limit threshold of the voltage (for example, from 340V to 310V), so as to avoid false triggering of protection. If the prediction result is "continuous running": the system maintains the normal running voltage protection threshold (such as lower limit 340V and upper limit 420V). If the prediction result is "planned shutdown": the system can adjust the upper and lower threshold values to more strict values to detect the voltage mutation in the abnormal shutdown process in advance. If the prediction result is "possible abnormality": the system can enable the redundant protection mechanism, increase the sampling frequency and narrow the protection threshold range (such as setting the upper and lower limits to 370V~390V), so as to realize more sensitive abnormal voltage detection.

[0104] In the process of intelligent voltage protection of the target device by the device voltage protection threshold, the dynamic voltage protection specifically comprises: the adjusted voltage protection threshold is issued to the voltage monitoring module and the control module in real time. When the actually collected voltage value exceeds the dynamically adjusted protection threshold range, the system triggers the corresponding intelligent protection action, for example, the device emergency operation that can be taken includes: sending an alarm signal; controlling the device to be powered off; pushing the voltage abnormal information to the platform monitoring center; starting the backup power supply or switching the power supply path (such as the system has UPS or dual power supply switching capability).

[0105] In the process of intelligent voltage protection of the target device by the device voltage protection threshold, the recovery and adaptive mechanism specifically comprises: after the protection action is completed, the system automatically restores the normal protection threshold setting according to the device recovery state, external behavior identification and new round of prediction results, to realize closed-loop control.

[0106] The application can significantly improve the adaptability of the system to voltage fluctuations by linking the future operation trend of the device with the voltage protection strategy, avoid false positives, false negatives and unnecessary downtime, and enhance the operation stability and intelligent protection level of the device under complex working conditions.

[0107] In addition, another aspect of the application, in some embodiments, the application provides an intelligent voltage protection system based on Internet of Things, which comprises a voltage protection unit, for reference Figure 2 The figure is a structural schematic diagram of an exemplary hardware and / or software of a voltage protection unit according to some embodiments of the application, which comprises a device action identification module 201, a decision module 202, a device behavior prediction module 203 and a voltage protection adjustment module 204, which are described as follows:

[0108] The device action identification module 201 is used to collect external behavior associated data of the target device through Internet of Things, to identify the device action according to the external behavior associated data, and to obtain the device external action identification result of the target device;

[0109] The device action identification module 201 is also used to collect internal voltage state data of the target device, to identify the device action according to the internal voltage state data, and to obtain the internal action identification result of the target device;

[0110] The decision module 202 is used to perform behavior consistency test according to the device external action identification result and the internal action identification result, to obtain the external behavior consistency degree;

[0111] The device behavior prediction module 203 is configured to acquire image sensing data of the target device when the external behavior consistency is higher than the preset trusted threshold, perform dynamic behavior prediction on the target device according to the image sensing data and the external behavior consistency, and determine a device behavior prediction result.

[0112] The voltage protection adjustment module 204 is configured to dynamically adjust a device voltage protection threshold according to the behavior prediction result, and perform intelligent voltage protection on the target device by using the device voltage protection threshold.

[0113] The above describes an example of the smart voltage protection system and method based on the Internet of Things provided by the embodiments of the present application in detail. It can be understood that the corresponding device contains the corresponding hardware structure and / or software module for executing each function in order to realize the above functions.

[0114] Those skilled in the art should easily realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function in the present application is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution, so that a person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0115] In addition, the present application also provides a computer terminal device, which comprises a memory and a processor, the memory stores a code, and the processor is configured to acquire the code and execute the above-mentioned smart voltage protection method based on the Internet of Things.

[0116] In some embodiments, with reference to Figure 3 The figure is a structural schematic diagram of a computer terminal device for implementing a smart voltage protection method based on the Internet of Things according to some embodiments of the present application. The smart voltage protection method based on the Internet of Things in the above-mentioned embodiments can be implemented by the computer terminal device shown in the figure, which comprises at least one communication bus 301, a communication interface 302, a processor 303, and a memory 304. Figure 3

[0117] The processor 303 can be a general central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more processors for controlling the execution of the smart voltage protection method based on the Internet of Things in the present application. ​

[0118] The communication bus 301 can include a path for communicating information among the above-mentioned components.

[0119] The memory 304 can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk storage or other magnetic storage devices, or any other medium capable of storing desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited to this. The memory 304 can exist independently, and is connected to the processor 303 through the communication bus 301. The memory 304 can also be integrated with the processor 303.

[0120] The memory 304 is configured to store program code for implementing the scheme of the present application, and the processor 303 is configured to control the execution. The processor 303 is configured to execute the program code stored in the memory 304. The program code can include one or more software modules. The determination of the consistency of the external behavior in the above-mentioned embodiments can be implemented by one or more software modules in the program code of the processor 303 and the memory 304.

[0121] The communication interface 302 uses any transceiver-like device for communicating with other devices or communication networks, such as an Ethernet, a radio access network (RAN), a wireless local area network (WLAN), etc.

[0122] Optionally, the computer terminal device 300 can further include a power supply 305 for providing power to various devices or circuits in the real-time computer terminal device.

[0123] In a specific implementation, as an embodiment, the computer terminal device can include a plurality of processors, each of which can be a single-CPU processor or a multi-CPU processor. The processor here can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0124] The computer terminal device described above can be a general-purpose computer terminal device or a special-purpose computer terminal device. In a specific implementation, the computer terminal device can be a desktop computer, a laptop computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of computer terminal device.

[0125] In addition, the other aspects of the present application also provide a computer-readable storage medium, which stores at least one computer program, and the computer program is loaded and executed by a processor to implement the operations performed by the above-mentioned smart voltage protection method based on Internet of Things.

[0126] In summary, in the smart voltage protection system and method based on Internet of Things disclosed by the embodiments of the present application, first, the external behavior correlation data of the target device is collected through Internet of Things, the device action recognition is performed according to the behavior correlation data, and the external action recognition result of the target device is obtained; the internal voltage state data of the target device is collected, the device action recognition is performed according to the internal voltage state data, and the internal action recognition result of the target device is obtained; the behavior consistency test is performed according to the external action recognition result and the internal action recognition result, and the external behavior consistency degree is obtained; when the external behavior consistency degree is higher than the preset trusted threshold, the image sensing data of the target device is acquired, the behavior dynamic prediction of the target device is performed according to the image sensing data and the external behavior consistency degree, and the device behavior prediction result is determined; the device voltage protection threshold is dynamically adjusted according to the behavior prediction result, and the smart voltage protection of the target device is performed by the device voltage protection threshold, which can perform the behavior prediction of the target device through the image sensing data of Internet of Things, and dynamically adjust the voltage protection threshold according to the behavior prediction result, thereby avoiding the voltage false alarm caused by frequent start and stop.

[0127] The above is only an embodiment of the present application, and the specific technical solutions or common knowledge of the scheme are not described in detail. It should be noted that, for those skilled in the art, without departing from the technical solutions of the present application, some modifications and improvements can be made, which should also be considered as the protection scope of the present application, and these will not affect the effect and practicality of the patent implementation.

[0128] The scope of the application for which protection is sought should be determined from the claims. The description, including the detailed description and drawings, is to be construed as illustrative only and not as restrictive. It will be obvious to those skilled in the art that various modifications and changes can be made to the application without departing from the spirit and scope thereof. Accordingly, it is intended that all such modifications and changes be included within the scope of the application as claimed.

Claims

1. An intelligent voltage protection method based on the Internet of Things, characterized in that: include: Collecting external behavior association data of the target device through the Internet of Things, performing device action recognition based on the external behavior association data, and obtaining a device external action recognition result of the target device; Collecting internal voltage state data of a target device, performing device action recognition based on the internal voltage state data, and obtaining an internal action recognition result of the target device; Performing a behavior consistency check based on the device's external action recognition result and the internal action recognition result to obtain an external behavior consistency; When the external behavior consistency is higher than a preset trustworthy threshold, acquiring image sensor data of the target device, performing dynamic behavior prediction on the target device based on the image sensor data and the external behavior consistency, and determining a device behavior prediction result; Dynamically adjusting a device voltage protection threshold according to the behavior prediction result, and performing intelligent voltage protection on the target device by the device voltage protection threshold; The step of dynamically predicting the behavior of the target device based on the image sensing data and the external behavior consistency and determining the device behavior prediction result specifically includes: After preprocessing the image sensing data, time series image features are extracted using a convolutional neural network; Determining a behavior prediction depth of the target device according to the external behavior consistency, performing moving average autoregressive modeling on the time series image features based on the behavior prediction depth, and determining a device behavior prediction result of the target device through the moving average autoregressive model; According to the range of the external behavior consistency, dynamic mapping is performed according to a preset mapping table to determine the number of image frames as the corresponding behavior prediction depth; The external behavior-related data includes: external image data, external sound data, and external vibration data of the target device, and the external behavior-related data is collected in real time through an Internet of Things sensor.

2. The method according to claim 1, wherein The internal voltage status data is collected by a voltage sensor installed in the power supply line loop of the target device.

3. The method according to claim 1, wherein Performing device action recognition based on the internal voltage state data to obtain an internal action recognition result of the target device specifically includes: Constructing a time series sample set of a voltage signal using a sliding time window according to the internal voltage state data; Extract voltage features from the voltage data in each time window, and construct voltage feature vectors based on multiple voltage features corresponding to each time window to obtain a feature vector set; Each voltage feature vector in the feature vector set is subjected to feature classification through a trained neural network to obtain an internal action state label corresponding to each time window, thereby obtaining an internal action recognition result.

4. The method according to claim 1, wherein Performing a behavior consistency check based on the device's external action recognition result and the internal action recognition result to obtain the external behavior consistency specifically includes: Perform code value mapping based on the device external action recognition result label corresponding to each time window to obtain the corresponding external action recognition code sequence; Map the code values ​​according to the device internal action recognition result labels corresponding to each time window to obtain the corresponding internal action recognition code sequence; A consistency check is performed based on the external action recognition coding sequence and the internal action recognition coding sequence to determine the external behavior consistency.

5. The method according to claim 1, wherein Performing device action recognition based on the external behavior association data to obtain a device external action recognition result of the target device specifically includes: Acquiring external image data, external sound data, and external vibration data from the behavior-related data; performing image action recognition on the external image data to determine a first action recognition result; performing device sound recognition on the external sound data to determine a second action recognition result; performing vibration abnormality recognition on the external vibration data to determine a third action recognition result; A fusion decision is performed based on the first, second, and third action recognition results to determine an external device action recognition result of the target device.

6. An intelligent voltage protection system based on the Internet of Things, comprising a voltage protection unit, wherein the voltage protection unit is configured to execute the intelligent voltage protection method based on the Internet of Things according to any one of claims 1 to 5, characterized in that: The voltage protection unit includes: A device action recognition module is used to collect external behavior association data of a target device through the Internet of Things, perform device action recognition based on the external behavior association data, and obtain a device external action recognition result of the target device; The device action recognition module is further configured to collect internal voltage status data of a target device, perform device action recognition based on the internal voltage status data, and obtain an internal action recognition result of the target device; a decision module, configured to perform a behavior consistency check based on the device's external action recognition result and the internal action recognition result to obtain an external behavior consistency; a device behavior prediction module, configured to, when the external behavior consistency is higher than a preset trustworthy threshold, obtain image sensor data of the target device, dynamically predict the behavior of the target device based on the image sensor data and the external behavior consistency, and determine a device behavior prediction result; The voltage protection adjustment module is used to dynamically adjust the device voltage protection threshold according to the behavior prediction result, and perform intelligent voltage protection on the target device by the device voltage protection threshold.

7. A computer terminal device, characterized in that: The computer terminal device includes a memory and a processor, the memory stores a code, and the processor is configured to obtain the code and execute the intelligent voltage protection method based on the Internet of Things as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing at least one computer program, characterized in that: The computer program is loaded and executed by a processor to implement the operations performed by the intelligent voltage protection method based on the Internet of Things as described in any one of claims 1 to 5.

Images