Sensor data front-end processing method, device, equipment, medium and program

Abstract

The invention relates to a sensor data front-end processing method and device, equipment, a storage medium and a computer program, and the method comprises the steps: collecting a real-time operation data set of a target sensor, and carrying out the data filtering of the real-time operation data set, and obtaining a target data set; constructing a data feature map of the target data set, and extracting target features of the target data set according to the data feature map; performing state analysis on preset monitoring equipment corresponding to the target sensor according to the target feature to obtain an operation state of the monitoring equipment; and intelligently adjusting the target sensor according to the running state. According to the invention, the data front-end processing capability of the sensor is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a sensor data front-end processing method, device, equipment, medium and program. Background Art

[0002] Generally speaking, traditional sensors are mainly responsible for collecting physical quantities, such as temperature, humidity, pressure, light intensity, etc., and converting them into electrical signals, such as voltage, current, etc. These electrical signals are then transmitted to other subsequent analog-to-digital conversion devices or instruments for processing, and then uploaded to the upper-level platform for further analysis. In this process, the function of the sensor itself is relatively single, focusing only on data collection and preliminary conversion. Further processing and interpretation of the data depend on external devices and upper-level servers.

[0003] The existing technology processes the data collected by sensors in external devices, requiring the entire system to include multiple components, such as sensors, data transmission lines, analog-to-digital conversion devices, external processors, and corresponding software systems. This approach results in longer data links and increases the difficulty of system design, installation, and maintenance. Especially in some remote monitoring or large-scale data acquisition systems, data transmission delays will be more obvious, and the processing of large amounts of data will also put greater pressure on server computing and storage resources. At the same time, sensors lack the ability to adapt to different application scenarios and cannot automatically adjust measurement parameters or use different algorithms to improve accuracy.

[0004] Therefore, how to improve the front-end processing capabilities of sensor data has become an urgent problem to be solved. Summary of the invention

[0005] The present application provides a sensor data front-end processing method, device, equipment, medium and program to solve the problem of poor sensor data front-end processing capability.

[0006] In a first aspect, the present application provides a sensor data front-end processing method, comprising:

[0007] Collecting a real-time operation data set of a target sensor, and performing data filtering on the real-time operation data set to obtain a target data set;

[0008] Constructing a data feature graph of the target data set, and extracting target features of the target data set according to the data feature graph;

[0009] Performing a status analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating status of the monitoring device;

[0010] The target sensor is intelligently adjusted according to the operating status.

[0011] In some embodiments, performing data filtering on the real-time running data set to obtain a target data set includes:

[0012] Get the size and standard deviation of the preset Gaussian kernel;

[0013] Generate a Gaussian kernel matrix according to the size of the Gaussian kernel and the standard deviation;

[0014] Filling edge pixels of the real-time running data set according to the Gaussian kernel matrix to obtain a filled data set;

[0015] Convolving the Gaussian kernel with the padded data set to obtain a plurality of pixel values;

[0016] The plurality of pixel values ​​are collected as a target data set.

[0017] In some embodiments, constructing a data feature graph of the target data set includes:

[0018] Dividing the target data set into data sources to obtain one-dimensional time series data;

[0019] Performing Fourier transform on the one-dimensional time series data to obtain one-dimensional frequency domain data;

[0020] A feature matrix is ​​constructed according to the one-dimensional frequency domain data to obtain a data feature map of the target data set.

[0021] In some embodiments, extracting the target features of the target data set according to the data feature graph includes:

[0022] Performing multi-layer convolution on the data feature map to obtain a spatial feature map;

[0023] Performing multi-layer self-attention calculation on the spatial feature map to obtain self-attention features at each level;

[0024] The self-attention features are fused to obtain target features.

[0025] In some embodiments, performing a status analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain the operating status of the monitoring device includes:

[0026] Performing feature encoding on the target feature to obtain a coded feature;

[0027] Performing activation and pooling processing on the coding features to obtain nonlinear features;

[0028] The nonlinear features are fully connected and mapped to obtain the operating status of the preset monitoring device corresponding to the target sensor.

[0029] In some embodiments, the intelligently adjusting the target sensor according to the operating state includes:

[0030] Invoking a sensor adjustment strategy corresponding to the operating state;

[0031] Calculating the similarity between the adjustment strategy and a pre-constructed adjustment state set;

[0032] A matching adjustment state is determined according to the similarity, and the target sensor is adjusted according to the matching adjustment state.

[0033] In a second aspect, the present application provides a sensor data front-end processing device, comprising:

[0034] A data filtering module is used to collect a real-time operation data set of a target sensor, and perform data filtering on the real-time operation data set to obtain a target data set;

[0035] A feature extraction module, used to construct a data feature graph of the target data set, and extract target features of the target data set according to the data feature graph;

[0036] A state analysis module, used to perform state analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating state of the monitoring device;

[0037] An intelligent adjustment module is used to intelligently adjust the target sensor according to the operating state.

[0038] In a third aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the above aspects.

[0039] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the above aspects when executed by a processor.

[0040] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the steps of the method described in the above aspects when executed by a processor.

[0041] The present application provides a sensor data front-end processing method, device, equipment, medium and program, which utilizes data filtering to denoise a real-time operation data set, and can significantly reduce the noise of the real-time operation data set, making the real-time operation data set look clearer and more natural, while retaining the main data features and details, and improving the data acquisition quality and visualization effect; by extracting the spatiotemporal features of the target data set, it is possible to learn the spatial and temporal feature relationships between the target data sets, and to more deeply explore the data features of the target data sets, and to improve the accuracy of subsequent operation status analysis; by deeply analyzing the target features, it is possible to more accurately capture the operation status changes of the preset monitoring equipment corresponding to the target sensor, which helps to timely discover potential problems and improve the response speed and efficiency of sensor processing; by matching with a pre-built adjustment state set, it is possible to ensure that the adjustment strategy adopted is verified and effective, thereby improving the accuracy of the adjustment, and at the same time, the pre-built adjustment state set can greatly shorten the time for finding and adjusting the strategy, and improve the efficiency and adaptability of sensor adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present application will be described in more detail below based on embodiments and with reference to the accompanying drawings:

[0043] Figure 1 A schematic diagram of a flow chart of a sensor data front-end processing method provided in an embodiment of the present application;

[0044] Figure 2 A schematic diagram of functional modules of a sensor data front-end processing device provided in an embodiment of the present application;

[0045] Figure 3 A schematic diagram of the structure of an electronic device for a sensor data front-end processing method provided in an embodiment of the present application.

[0046] In the drawings, the same reference numerals are used for the same components, and the drawings are not drawn to scale. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the technical solution of the present application, and to fully understand and implement how the present application applies technical means to solve technical problems and achieve the corresponding technical effects, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only embodiments of a part of the present application, not all embodiments. The embodiments of the present application and the various features in the embodiments can be combined with each other without conflict, and the technical solutions formed are all within the scope of protection of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work should belong to the scope of protection of the present application.

[0048] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0049] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0050] The embodiment of the present application provides a sensor data front-end processing method. The execution subject of the sensor data front-end processing method includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the system provided by the embodiment of the present application. In other words, the sensor data front-end processing method can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

[0051] Embodiment 1

[0052] Figure 1 A schematic diagram of a flow chart of a sensor data front-end processing method provided in an embodiment of the present application, such as Figure 1 As shown, the sensor data front-end processing method includes:

[0053] S1. Collect a real-time operation data set of a target sensor, and perform data filtering on the real-time operation data set to obtain a target data set.

[0054] In an embodiment of the present invention, the real-time operation data set is real-time operation data collected by the corresponding target sensor selected according to the data type to be collected. For example, when the data type to be collected is temperature and humidity data, a temperature sensor is selected to collect relevant data, such as collecting the current ambient temperature and humidity, etc., and the target sensor is calibrated before use to ensure the accuracy of its measurement results. The calibration process includes adjusting the sensor's zero point, sensitivity and other parameters.

[0055] In the embodiment of the present invention, the step of filtering the real-time running data set to obtain the target data set includes:

[0056] Get the size and standard deviation of the preset Gaussian kernel;

[0057] Generate a Gaussian kernel matrix according to the size of the Gaussian kernel and the standard deviation;

[0058] Filling edge pixels of the real-time running data set according to the Gaussian kernel matrix to obtain a filled data set;

[0059] Convolving the Gaussian kernel with the padded data set to obtain a plurality of pixel values;

[0060] The plurality of pixel values ​​are collected as a target data set.

[0061] Wherein, the data filtering may include Gaussian filtering, median filtering, etc., which are not specifically limited here. For example, the present invention may utilize Gaussian filtering to denoise the real-time running data set, wherein the size of the Gaussian kernel depends on the degree of smoothness required after denoising the real-time running data set and the size of the real-time running data set, and the standard deviation of the Gaussian kernel affects the degree of smoothness after denoising the real-time running data set. The larger the standard deviation of the Gaussian kernel, the better the smoothing effect. Before Gaussian filtering is performed on the real-time running data set, the edge pixels of the real-time running data set need to be filled to ensure that the data set has complete edge pixels so that the Gaussian kernel can accurately cover all parts of the real-time running data set.

[0062] Furthermore, the convolution operation is to traverse the Gaussian kernel through the padded data set, each time covering the kernel on a local area of ​​the data set, making the central element of the Gaussian kernel matrix correspond to the central pixel of the padded data set, and then performing weighted averaging on the Gaussian kernel matrix and the corresponding pixels in the padded data set, and finally adding the weighted average results to obtain several pixel values, namely the convolution results, and the several pixel values ​​are collected to obtain the target data set.

[0063] In the embodiment of the present invention, data filtering is used to denoise the real-time running data set. By using the method of weighted averaging of surrounding pixels, the noise of the real-time running data set can be significantly reduced, making the real-time running data set look clearer and more natural, while retaining the main data features and details, thereby improving the data collection quality and visualization effect.

[0064] S2. Construct a data feature graph of the target data set, and extract target features of the target data set according to the data feature graph.

[0065] In an embodiment of the present invention, the data feature map reconstructs a one-dimensional target data set into a two-dimensional feature map structure, and then the spatial relationship between different data in the target data set can be captured according to the two-dimensional feature map to improve the accuracy and reliability of extracting target features.

[0066] In an embodiment of the present invention, the step of constructing a data feature graph of the target data set includes:

[0067] Dividing the target data set into data sources to obtain one-dimensional time series data;

[0068] Performing Fourier transform on the one-dimensional time series data to obtain one-dimensional frequency domain data;

[0069] A feature matrix is ​​constructed according to the one-dimensional frequency domain data to obtain a data feature map of the target data set.

[0070] In an embodiment of the present invention, the target data set includes real-time operation data of the data source corresponding to each target sensor. Therefore, the target data set can be divided into data source data corresponding to each target sensor, and the data source data is sorted according to time points to obtain one-dimensional time series data. The one-dimensional time series data is Fourier transformed to convert it from the time domain to the frequency domain. Each one-dimensional frequency domain data is used as a row vector to construct a feature matrix to obtain a data feature map characterizing the distribution of the data source data, so that the features in the frequency domain can be extracted, that is, the target features can be obtained.

[0071] In the embodiment of the present invention, extracting the target features of the target data set according to the data feature graph includes:

[0072] Performing multi-layer convolution on the data feature map to obtain a spatial feature map;

[0073] Performing multi-layer self-attention calculation on the spatial feature map to obtain self-attention features at each level;

[0074] The self-attention features are fused to obtain target features.

[0075] In detail, the pre-built two-layer convolutional layer can be used to extract the spatial features of the data feature map, and the LSTM unit can be used to extract and retain the temporal features to obtain the self-attention features containing temporal and spatial features, and the self-attention calculations are performed separately through two independent LSTM units to obtain self-attention features of different levels, and the self-attention features of different levels are feature fused to more accurately extract the temporal dependency in the spatial feature map and obtain more accurate spatiotemporal features of the target data set.

[0076] Furthermore, feature fusion of self-attention features at different levels can be performed by scaling the single feature maps corresponding to self-attention features at different levels through a preset target spatial resolution, and splicing the scaled feature maps. The data features on each pixel in the spliced ​​feature map are the obtained target features, wherein the target spatial resolution is usually a fixed size, such as width and height, and the single feature maps corresponding to self-attention features at different levels of the target spatial resolution are converted into feature maps within the target spatial resolution by spatial upsampling or downsampling, thereby improving the accuracy of subsequent target sensor state analysis.

[0077] In the embodiment of the present invention, by extracting the spatiotemporal features of the target data set, the spatial and temporal feature relationships between the target data sets can be learned, the data features of the target data sets can be mined at a deeper level, and the accuracy of subsequent operation status analysis can be improved.

[0078] S3. Performing status analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating status of the monitoring device.

[0079] In an embodiment of the present invention, the state analysis of the target sensor is performed according to the target feature, that is, the target feature is encoded to convert the original feature data into a format suitable for processing by a preset neural network, and then fully connected mapping is performed to obtain the feature space corresponding to each operating state, thereby obtaining the operating state of the preset monitoring device corresponding to the target sensor.

[0080] In the embodiment of the present invention, the step of performing a state analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain the operating state of the monitoring device includes:

[0081] Performing feature encoding on the target feature to obtain a coded feature;

[0082] Performing activation and pooling processing on the coding features to obtain nonlinear features;

[0083] The nonlinear features are fully connected and mapped to obtain the operating status of the preset monitoring device corresponding to the target sensor.

[0084] Among them, the monitoring equipment can be industrial machinery equipment, power equipment, energy equipment, etc. For example, in vibration monitoring such as motors, pumps, fans, compressors, etc., by monitoring target characteristics such as vibration frequency and amplitude, the operating conditions and fault diagnosis such as bearing wear, rotor imbalance, and shaft misalignment can be analyzed to facilitate subsequent intelligent adjustment of target sensors.

[0085] Specifically, the spatiotemporal features can be encoded again through an independent pre-built LSTM (Long Short-Term Memory) network, which can alleviate the forgetting phenomenon and extract important temporal features to obtain more accurate encoding features.

[0086] Furthermore, the encoded features are used as inputs of the input layer in a preset neural network, activated and pooled to obtain nonlinear features, and one or more fully connected layers (hidden layers) are used to extract features, each fully connected layer contains multiple neurons, and each neuron is connected to all the features of the input layer. The fully connected layer constructed by two convolution modules is fully connected to the nonlinear features, and the nonlinear features are mapped to the feature space corresponding to each operating state to obtain the corresponding operating state of the monitoring equipment, wherein the operating state of the monitoring equipment can be a fault level such as normal, minor fault, severe fault and other operating states.

[0087] In the embodiment of the present invention, by deeply analyzing the target characteristics, the operating status changes of the monitoring equipment corresponding to the target sensor can be captured more accurately, which helps to discover potential problems in a timely manner, thereby avoiding misdiagnosis or missed diagnosis; at the same time, the status analysis can monitor the operating status in real time, greatly shortening the time from the occurrence of the problem to its discovery and processing, and improving the response speed and efficiency of sensor processing.

[0088] S4. Intelligently adjust the target sensor according to the operating status.

[0089] In an embodiment of the present invention, the adjustment strategy to be adopted is determined according to the operating status of the preset monitoring device corresponding to the target sensor (such as normal, minor fault, serious fault, etc.), and the adjustment strategy includes adjusting the sensitivity, range, zero drift compensation and other parameters of the sensor. The target sensor of the corresponding operating status is intelligently adjusted according to the adjustment strategy.

[0090] In an embodiment of the present invention, the intelligently adjusting the target sensor according to the operating state includes:

[0091] Invoking a sensor adjustment strategy corresponding to the operating state;

[0092] Calculating the similarity between the adjustment strategy and a pre-constructed adjustment state set;

[0093] A matching adjustment state is determined according to the similarity, and the target sensor is adjusted according to the matching adjustment state.

[0094] In detail, calling the sensor adjustment strategy corresponding to the operating state refers to determining the sensor adjustment strategy corresponding to the target sensor setting corresponding to different operating states, the sensor adjustment strategy includes sensor zero point adjustment, sensitivity adjustment, linearity adjustment, etc., and the adjustment state set is the operating data corresponding to the target sensor under different adjustment strategies. For example, after the sensor zero point adjustment, the output data of the sensor near the zero point is adjusted more finely through the adjustment strategy, and then the target sensor is adjusted. The similarity between the adjustment strategy and the pre-built adjustment state set can be calculated by calculating the feature similarity to match an adjustment strategy that is more suitable for the target sensor. For example, after determining the sensor adjustment strategy corresponding to the operating state, the similarity between the operating data output by the target sensor after adjustment is calculated and the operating data of the target sensor in the adjustment state set is calculated.

[0095] Furthermore, the target sensor is adjusted accordingly according to the adjustment strategy corresponding to the matched adjustment state, including adjusting the physical settings of the sensor, sending a control instruction to the sensor to change its working mode or parameters, etc.

[0096] In an embodiment of the present invention, by matching with a pre-built adjustment state set, it can be ensured that the adjustment strategy adopted is verified and effective, thereby improving the accuracy of the adjustment. At the same time, the pre-built adjustment state set can greatly shorten the time for finding and adjusting the strategy, and improve the efficiency and adaptability of sensor adjustment.

[0097] The present invention utilizes data filtering to denoise the real-time operation data set, which can significantly reduce the noise of the real-time operation data set, making the real-time operation data set look clearer and more natural, while retaining the main data features and details, and improving the data acquisition quality and visualization effect; by extracting the spatiotemporal features of the target data set, the spatial and temporal feature relationships between the target data sets can be learned, and the data features of the target data set can be mined at a deeper level to improve the accuracy of subsequent operation status analysis; by deeply analyzing the target features, the operation status changes of the preset monitoring equipment corresponding to the target sensor can be more accurately captured, which helps to discover potential problems in a timely manner and improve the response speed and efficiency of sensor processing; by matching with the pre-constructed adjustment state set, it can ensure that the adjustment strategy adopted is verified and effective, thereby improving the accuracy of the adjustment. At the same time, the pre-constructed adjustment state set can greatly shorten the time for finding and adjusting the strategy, and improve the efficiency and adaptability of sensor adjustment.

[0098] Embodiment 2

[0099] like Figure 2 , which is a functional module diagram of a sensor data front-end processing device 100 provided for this embodiment.

[0100] The sensor data front-end processing device 100 described in the present invention can be installed in an electronic device. According to the functions to be implemented, the sensor data front-end processing device 100 can include a data filtering module 101, a feature extraction module 102, a state analysis module 103, and an intelligent adjustment module 104. The module described in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0101] In this embodiment, the functions of each module / unit are as follows:

[0102] The data filtering module 101 is used to collect a real-time operation data set of a target sensor, and perform data filtering on the real-time operation data set to obtain a target data set;

[0103] A feature extraction module 102 is used to construct a data feature graph of the target data set, and extract target features of the target data set according to the data feature graph;

[0104] A state analysis module 103 is used to perform state analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating state of the monitoring device;

[0105] The intelligent adjustment module 104 is used to intelligently adjust the target sensor according to the operating state.

[0106] Embodiment 3

[0107] Figure 3 A schematic diagram of the structure of an electronic device for a sensor data front-end processing method provided in an embodiment of the present application.

[0108] On the basis of the above embodiments, this embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the above embodiments.

[0109] In some implementations of this embodiment, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0110] In some implementations of this embodiment, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0111] The processor may include, but is not limited to, one or more processors or microprocessors, etc. Each processor may be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components to execute the methods in the above embodiments.

[0112] The computer-readable storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, and the computer-readable storage medium may include but is not limited to, for example, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disk, floppy disk, solid-state drive, removable disk, CD ROM, DVD ROM, Blu-ray disc, etc.).

[0113] The computer-readable storage medium may also store at least one computer executable program, such as a computer-readable instruction. The computer-readable storage medium includes, but is not limited to, for example, a volatile memory and / or a non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The computer-readable storage medium may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device runs the computer-readable instructions stored on the computer-readable storage medium, the various methods described above may be performed.

[0114] In addition, the computer device may also include (but not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (eg, keyboard, mouse, speaker, etc.), etc.

[0115] The processor may communicate with a communication interface of an external device via an I / O bus via a wired or wireless network.

[0116] In one embodiment, the at least one computer executable instruction may also be compiled into or constitute a software product / computer program product, wherein one or more computer executable instructions are executed by a processor to perform the various functions and / or method steps in the embodiments described in the present technology.

[0117] In the embodiments provided in the present application, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the above-mentioned module, a program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0118] It should be noted that, in this application, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element limited by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0119] Although the implementation methods disclosed in this application are as above, the above contents are only the implementation methods adopted for facilitating the understanding of this application, and are not intended to limit this application. Any technician in the technical field to which this application belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in this application, but the scope of patent protection of this application shall still be based on the scope defined in the attached claims.

Claims

1. A sensor data front-end processing method, characterized in that: The method comprises: Collecting a real-time operation data set of a target sensor, and performing data filtering on the real-time operation data set to obtain a target data set; Constructing a data feature graph of the target data set, and extracting target features of the target data set according to the data feature graph; Performing a status analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating status of the monitoring device; The target sensor is intelligently adjusted according to the operating status.

2. A sensor data front-end processing method according to claim 1, characterized in that: The performing data filtering on the real-time running data set to obtain a target data set includes: Get the size and standard deviation of the preset Gaussian kernel; Generate a Gaussian kernel matrix according to the size of the Gaussian kernel and the standard deviation; Filling edge pixels of the real-time running data set according to the Gaussian kernel matrix to obtain a filled data set; Convolving the Gaussian kernel with the padded data set to obtain a plurality of pixel values; The plurality of pixel values ​​are collected as a target data set.

3. A sensor data front-end processing method according to claim 1, characterized in that: The step of constructing a data feature graph of the target data set includes: Dividing the target data set into data sources to obtain one-dimensional time series data; Performing Fourier transform on the one-dimensional time series data to obtain one-dimensional frequency domain data; A feature matrix is ​​constructed according to the one-dimensional frequency domain data to obtain a data feature map of the target data set.

4. A sensor data front-end processing method according to claim 3, characterized in that: The step of extracting the target feature of the target data set according to the data feature graph comprises: Performing multi-layer convolution on the data feature map to obtain a spatial feature map; Performing multi-layer self-attention calculation on the spatial feature map to obtain self-attention features at each level; The self-attention features are fused to obtain target features.

5. The sensor data front-end processing method according to claim 1, characterized in that: The performing of state analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain the operating state of the monitoring device includes: Performing feature encoding on the target feature to obtain a coded feature; Performing activation and pooling processing on the coding features to obtain nonlinear features; The nonlinear features are fully connected and mapped to obtain the operating status of the preset monitoring device corresponding to the target sensor.

6. A sensor data front-end processing method according to claim 1, characterized in that: The intelligently adjusting the target sensor according to the operating state includes: Invoking a sensor adjustment strategy corresponding to the operating state; Calculating the similarity between the adjustment strategy and a pre-constructed adjustment state set; A matching adjustment state is determined according to the similarity, and the target sensor is adjusted according to the matching adjustment state.

7. A sensor data front-end processing device, characterized in that: The device comprises: A data filtering module is used to collect a real-time operation data set of a target sensor, and perform data filtering on the real-time operation data set to obtain a target data set; A feature extraction module, used to construct a data feature graph of the target data set, and extract target features of the target data set according to the data feature graph; A state analysis module, used to perform state analysis on a preset monitoring device corresponding to the target sensor according to the target feature to obtain an operating state of the monitoring device; An intelligent adjustment module is used to intelligently adjust the target sensor according to the operating state.

8. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images