A full-performance detection data processing method and device based on sensor fusion and a medium

Abstract

This invention discloses a method, device, and medium for processing full-performance detection data based on sensor fusion, belonging to the field of data processing technology. It includes preprocessing multi-sensor detection data using an adaptive spatiotemporal correlation denoising and repair algorithm to obtain preprocessed detection data; using a spatiotemporal correlation matrix for anomaly detection, identifying outliers in the sensor data, and performing repair and denoising; and intelligently enhancing the preprocessed detection data using an adaptive temporal multidimensional information enhancement and intelligent fusion algorithm to obtain enhanced detection data. This invention accurately detects and repairs abnormal sensor data through dynamic thresholds and incremental information transmission. Combined with an intelligent fusion algorithm, it dynamically adjusts sensor weights and gains to optimize data fusion accuracy. It maximizes information entropy to enhance adaptability, effectively responding to environmental interference, improving data reliability and accuracy, and providing high-quality support for subsequent analysis.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method, device, and medium for processing full-performance detection data based on sensor fusion. Background Technology

[0002] In full-performance laboratories, the application of sensor fusion technology presents unique challenges and requirements. Full-performance laboratories are primarily used to evaluate and validate the performance of complex systems under various operating conditions. They typically involve multiple types of sensors to monitor and collect various physical quantities of the system, such as temperature, pressure, vibration, and humidity. The combination of these sensors helps provide comprehensive data on system operation, but it also introduces technical challenges in data processing and fusion.

[0003] In a full-performance laboratory, multiple sensors are often deployed in different test scenarios and experimental equipment. Each sensor has different measurement accuracy and characteristics, and its data is affected by factors such as environmental changes, equipment status, and sensor location. This diversity of data sources makes data fusion particularly complex. Especially when the experimental environment involves measurements with high precision requirements, errors, delays, and signal interference between sensors can significantly affect the experimental results, thereby impacting the final test and verification outcomes.

[0004] However, the above-mentioned technologies have at least the following technical problems: low accuracy and poor integrity in the processing of test data during the full-performance test data processing. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is to solve the problem of low accuracy and poor integrity of the detection data processing in the existing full-performance detection data processing process.

[0007] To address the aforementioned technical problems, this invention provides the following technical solution: a method for processing full-performance detection data based on sensor fusion, comprising the following steps:

[0008] An adaptive spatiotemporal correlation denoising and repair algorithm is used to preprocess the multi-sensor detection data to obtain preprocessed detection data. Anomaly detection is performed using the spatiotemporal correlation matrix to identify outliers in the sensor data and then repair and denoise them. The preprocessed detection data is then intelligently enhanced using an adaptive temporal multidimensional information enhancement and intelligent fusion algorithm to obtain enhanced detection data.

[0009] As a preferred embodiment of the full-performance detection data processing method based on sensor fusion described in this invention, the preprocessing includes: performing spatiotemporal feature decomposition and standardization on multi-source sensor data;

[0010] Acquire data from different sensors, and store the data from each sensor. At the point of time Output measurement data, represented as ,in, It is the dimension of the data output by the sensor each time, and any element is used as... This indicates that it is the first The o-th data from a sensor, o = 1, 2, ..., m;

[0011] Spatiotemporal feature decomposition is performed on sensor data to extract spatial and temporal features and identify patterns of change in the data. The spatiotemporal feature decomposition is expressed as follows:

[0012] ;

[0013] in, It is the first Each sensor at a time point Spatiotemporal feature decomposition representation on, Indicates the first Each sensor at a time point Spatial features on, It is the first Each sensor at a time point The time characteristics on;

[0014] Further analysis of arbitrary sensor data Normalization is performed to eliminate the influence of different dimensions from different sensors, resulting in standardized sensor data. .

[0015] As a preferred embodiment of the full-performance detection data processing method based on sensor fusion described in this invention, the anomaly detection includes: spatiotemporal feature decomposition representation based on sensor data. Establish a spatiotemporal correlation matrix Spatiotemporal correlation matrix Any element in the, i.e., the sensor With sensors At any moment The correlation is represented as:

[0016] ;

[0017] in, , and , They represent the first Spatial and temporal characteristics of each sensor It is the first adjustment parameter. It is the second adjustment parameter; Indicates sensor and sensors Between at a certain point in time Similarity on;

[0018] Using the spatiotemporal correlation matrix Define dynamic threshold It detects outliers in sensor data. When the difference between the measured value of any sensor and the mean value of the corresponding spatiotemporal correlation neighborhood exceeds a dynamic threshold, the current data point is considered to be outlier.

[0019] As a preferred embodiment of the full-performance detection data processing method based on sensor fusion described in this invention, the repair and denoising includes: processing the standardized sensor data... With dynamic threshold Comparison, If so, it is considered abnormal data. It is the average of the measured values ​​obtained by sensor i;

[0020] When an outlier is detected, an incremental information transmission mechanism is used for repair. This mechanism uses information from adjacent sensors to correct the outlier data. It is obtained by weighting the data differences with those of neighboring sensors;

[0021] The existing weighted mean filter is used to remove noise, resulting in denoised detection data. .

[0022] As a preferred embodiment of the sensor fusion-based full-performance detection data processing method described in this invention, the intelligent enhancement processing includes: based on the preprocessed detection data, using a nonlinear autoregressive model to establish a nonlinear time-series model to predict the detection data at the current moment. ;

[0023] Calculate the entropy value of each sensor's data. The formula for measuring the diversity and uncertainty of information is:

[0024] ;

[0025] in, Let M represent the probability value of the data distribution of sensor i at time t, where M is the number of different states or values. It is a variable index; the higher the entropy value, the more information the sensor data contains;

[0026] Based on entropy values, for each sensor At any moment Assign a dynamic weight The weight calculation formula is:

[0027] ;

[0028] in, For sensors At any moment The dynamic weights are defined as follows: n is the number of sensors, and i is the variable index.

[0029] As a preferred embodiment of the full-performance detection data processing method based on sensor fusion described in this invention, the intelligent enhancement processing further includes introducing an adaptive gain function to dynamically adjust the gain of each sensor data.

[0030] ;

[0031] in, It is a sensor The gain function of the sensor At any moment The gain value represents the adjustment factor for the sensor data; It is the first one after preprocessing Each sensor at time The mean, It is a sensor At any moment Standard deviation of data within a given time window It is a regularization factor.

[0032] As a preferred embodiment of the full-performance detection data processing method based on sensor fusion described in this invention, the intelligent enhancement processing further includes optimizing the fused and enhanced data based on an adaptive gain function. The enhanced detection data is specifically formulated as follows:

[0033] ;

[0034] in, It is enhanced detection data, which provides a foundation for subsequent data analysis.

[0035] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the aforementioned full-performance detection data processing method based on sensor fusion.

[0036] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the aforementioned full-performance detection data processing method based on sensor fusion.

[0037] The beneficial effects of this invention are as follows: By introducing an adaptive spatiotemporal correlation denoising and repair algorithm, this invention can dynamically denoise and repair outliers in multi-sensor data. Through spatiotemporal correlation modeling and incremental information transmission mechanisms, the algorithm effectively recovers the true signals in the sensor data, maintains the spatiotemporal consistency of the data, and improves the reliability and accuracy of the data.

[0038] By utilizing dynamic threshold adaptive adjustment and combining it with a spatiotemporal correlation matrix, anomalies in sensor data are accurately detected. Based on an incremental information transmission repair mechanism, not only is the influence of neighboring sensors considered, but also the accurate repair of abnormal data is achieved, avoiding data loss or erroneous repair.

[0039] The adaptive temporal multidimensional information enhancement and intelligent fusion algorithm dynamically adjusts the fusion process based on the weights and information content of different sensors, ensuring accurate fusion of data from multiple sensors in complex environments. This results in higher-quality enhanced data, providing robust data support for subsequent analysis. Furthermore, the introduction of information entropy maximization and an adaptive gain function dynamically adjusts sensor weights and gains according to data characteristics, improving the accuracy and stability of data fusion. This adaptive mechanism effectively addresses environmental complexity and differences in sensor characteristics, enhancing the flexibility and accuracy of data fusion. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 The above is a flowchart of a full-performance detection data processing method based on sensor fusion, which is provided as an embodiment of the present invention. Detailed Implementation

[0042] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0043] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a full-performance detection data processing method based on sensor fusion, including:

[0044] S1. The multi-sensor detection data is preprocessed using an adaptive spatiotemporal correlation denoising and repair algorithm to obtain the preprocessed detection data.

[0045] It should be noted that the introduction of the adaptive spatiotemporal correlation denoising and repair algorithm, by introducing spatiotemporal correlation modeling and incremental information transmission mechanism, can effectively recover the real signal from noise pollution and abnormal data, and maintain the spatiotemporal consistency of the data during the repair process.

[0046] Specifically, spatiotemporal feature decomposition and standardization are performed on multi-source sensor data; data from different sensors are acquired, and each sensor's data is processed... At the point of time Output a series of measurement data, represented as Where m is the dimension of the data output by the sensor each time, and any element is represented by... This indicates that it is the first The o-th data from a sensor, o = 1, 2, ..., m.

[0047] To perform unified processing of data from multiple sensors, spatiotemporal feature decomposition is required to enable the sensor data to reflect both spatial and temporal variations. This further identifies patterns of change in the data and provides clearer features for subsequent outlier detection. The spatiotemporal feature decomposition of any sensor data is represented as follows:

[0048] ;

[0049] in, It is the first Each sensor at a time point Spatiotemporal feature decomposition representation on; Indicates the first Each sensor at a time point Spatial characteristics, that is, the spatial distribution characteristics of the sensor at a certain moment, are used to capture sensor data in a certain spatial area using existing feature engineering techniques, such as the physical location of the sensor and the measurement range. It is the first Each sensor at a time point The temporal characteristics of the sensor data reflect its changing trend over time, containing time-series information about the sensor and describing how it responds to changes in the external environment over time. The purpose of the above decomposition steps is to extract the spatial and temporal characteristics from the sensor data, enabling subsequent processing to more accurately capture spatiotemporal correlations.

[0050] Further analysis of arbitrary sensor data Normalization is performed to eliminate the influence of different sensors due to their different dimensions, resulting in standardized sensor data. .

[0051] S2. Use the spatiotemporal correlation matrix to perform anomaly detection, identify outliers in sensor data, and perform repair and noise reduction.

[0052] Specifically, spatiotemporal feature decomposition representation based on sensor data Establish a spatiotemporal correlation matrix This allows for the quantification of the spatiotemporal dependencies between different sensors, providing a basis for anomaly detection.

[0053] Spatiotemporal correlation matrix Any element in the, i.e., the sensor With sensors At any moment The correlation can be expressed by the following formula:

[0054] ;

[0055] in, , and , They represent the first Spatial and temporal characteristics of each sensor; It is the first adjustment parameter, which is used to adjust the relative influence between time features. It affects the contribution of time features in spatiotemporal correlation and is obtained based on expert experience. It is the second adjustment parameter, which is the coefficient that adjusts the influence of the time component in the comprehensive measurement of spatial and temporal characteristics on the overall correlation measurement. It is combined with the spatial characteristics in the denominator to ensure the balance of the measurement values; it is obtained based on expert experience.

[0056] The above formula provides a measure of spatiotemporal correlation. This reflects the sensor and sensors Between at a certain point in time Similarity on the surface.

[0057] Furthermore, using the spatiotemporal correlation matrix Define a dynamic threshold This threshold is used to detect outliers in sensor data. A data point is considered outlier when the difference between a sensor's measurement and the mean of its spatiotemporally correlated neighborhood exceeds this threshold. Threshold As the correlation between sensors changes dynamically, it can adapt to anomaly detection in different scenarios:

[0058] ;

[0059] in, This is the anomaly tolerance coefficient, used to control the tolerance for outliers, and is obtained through experimental methods. For the number of sensors, These are elements in the spatiotemporal correlation matrix, representing sensors. and sensors The spatiotemporal correlation.

[0060] A global anomaly detection threshold is obtained by calculating the spatiotemporal correlation between all sensors and taking the average value.

[0061] Furthermore, the standardized sensor data With dynamic threshold Comparison, If so, it is considered abnormal data. It is the average value of the measured values ​​obtained by sensor i. Each sensor only collects one variable parameter. In this embodiment, the multi-source sensors include, but are not limited to, temperature, pressure, and humidity sensors, and the corresponding variable parameters are temperature, pressure, and humidity, respectively. The average value also refers to the average temperature, average pressure, average humidity, etc.

[0062] It should be noted that once an anomaly is detected, an incremental information transmission mechanism is used for repair.

[0063] The core idea of ​​incremental information transmission is to repair abnormal data by using information from adjacent sensors based on the spatiotemporal correlation between sensors.

[0064] Repaired data It is obtained by weighting the data differences with those of neighboring sensors.

[0065] The formula for incremental information transmission is as follows:

[0066] ;

[0067] in, This is the repaired data, representing the data of the i-th sensor at time t after being repaired through the incremental information transmission mechanism; It is the set of neighboring sensors of the i-th sensor, which includes other sensors that have strong spatiotemporal correlation with the i-th sensor in space or time; It represents the i-th sensor in the neighborhood of the i-th sensor. One sensor; It is the first Standardized data from each sensor; It is the standardized data from the i-th sensor; Incremental repair weights represent the sensor's... The impact weights on sensor i repair are calculated using the spatiotemporal correlation matrix:

[0068] ;

[0069] It is described that the repair weight is proportional to the spatiotemporal correlation between sensors, and the greater the correlation, the greater the contribution of the sensor to the repair result; It represents the k-th sensor in the neighborhood of the i-th sensor; and These represent the sensors. and sensors Spatiotemporal correlation, sensor and sensors The spatiotemporal correlation.

[0070] In this way, the repair results can better reflect the similarity between sensors and effectively remove outliers.

[0071] Furthermore, existing weighted mean filtering is applied to remove noise, ensuring that the repaired data is smoother and reducing the interference of noise on the results, thus obtaining denoised detection data. .

[0072] In summary, the preprocessed detection data obtained after the above treatment is obtained. .

[0073] S3. The preprocessed detection data is intelligently enhanced using an adaptive temporal multidimensional information enhancement and intelligent fusion algorithm to obtain enhanced detection data.

[0074] It should be noted that the preprocessed detection data is intelligently enhanced using an adaptive temporal multidimensional information enhancement and intelligent fusion algorithm to obtain enhanced detection data. The adaptive temporal multidimensional information enhancement and intelligent fusion algorithm, through temporal modeling and information fusion mechanisms, combined with complex adaptive adjustment and entropy optimization, enables sensor data from different sources to be intelligently enhanced and accurately fused in complex environments, ultimately outputting reliable and accurate data.

[0075] Specifically, based on the preprocessed detection data, a nonlinear time series model is established to better capture the dynamic characteristics of the preprocessed detection data.

[0076] Since the temporal dependence of the preprocessed detection data (i.e., sensor data) is non-linear, a non-linear autoregressive model is used. This model predicts the detection data at the current moment based on past data and external inputs (historical values ​​of sensor data). .

[0077] Furthermore, in order to dynamically adjust the fusion weights of each sensor, an information entropy maximization mechanism is introduced. Information entropy maximization is a principle used to measure the amount of information.

[0078] During the multi-sensor data fusion process, the weight of each sensor is dynamically adjusted so that the contribution of each sensor in the fusion is determined by the amount of information it contains.

[0079] Calculate the entropy value of each sensor's data. Entropy is used to measure the diversity and uncertainty of information; the formula for calculating entropy is:

[0080] ;

[0081] in, The probability value of the data distribution of sensor i at time t is obtained by performing frequency statistics on the data or by using estimation methods (such as maximum likelihood estimation); M is the number of different states or values ​​that may be obtained, that is, it defines the degree of discretization or number of categories of the data, which is determined according to the complexity of the data distribution and the pre-set discretization interval. It is a variable index.

[0082] The higher the entropy value, the more information the sensor data contains.

[0083] Based on entropy values, for each sensor At any moment Assign a dynamic weight The weight calculation formula is:

[0084] ;

[0085] in, For sensors At any moment The dynamic weighting allows sensors with large amounts of information to occupy a larger proportion in data fusion, thereby improving the quality of the final fused data.

[0086] Furthermore, an adaptive gain function is introduced to dynamically adjust the gain of each sensor's data, ensuring that the contribution of each sensor is proportional to its actual information content.

[0087] The adaptive gain function is defined as follows:

[0088] ;

[0089] in, It is a sensor The gain function of the sensor At any moment The gain value represents the adjustment factor for the sensor data; It is the first one after preprocessing Each sensor at time The mean; It is a sensor At any moment The standard deviation of data within a given time window reflects the degree of dispersion of the data; It is a regularization factor that controls the smoothness of the gain function, avoiding excessively large or small gain values ​​due to abnormal fluctuations in the data. It is determined based on expert experience.

[0090] The gain function is used to adjust the influence of each sensor on the final fusion result, so as to avoid the fusion result being overly affected by the existence of outliers in the data.

[0091] Furthermore, based on the adaptive gain function, the fused and enhanced data are optimized.

[0092] The enhanced detection data is expressed in the following formula:

[0093] ;

[0094] in, This is the enhanced detection data, which provides a foundation for subsequent data analysis.

[0095] Example 2 is the second embodiment of the present invention, which provides a full-performance detection data processing method based on sensor fusion. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0096] ① Multi-sensor data preprocessing based on adaptive spatiotemporal correlation denoising and repair algorithm

[0097] Step 1: Data Acquisition and Preprocessing

[0098] In practical applications, multiple sensors are used to collect different types of data, such as temperature, humidity, and vibration. In this embodiment, multiple sensors are used, such as sensor 1, sensor 2, sensor 3, etc., each at different time points. Output a series of measurement sensor data First, the measurement data undergoes spatiotemporal feature decomposition and standardization. The purpose of spatiotemporal feature decomposition is to extract the spatial and temporal features of each sensor and normalize the sensor data to eliminate the influence of different sensor data dimensions.

[0099] Step 2: Construction of the spatiotemporal correlation matrix

[0100] Based on the decomposed spatiotemporal feature data, a spatiotemporal correlation matrix is ​​established. This matrix reflects the time points between the sensors. The spatiotemporal correlation is used to quantify the similarity between sensors by calculating the spatiotemporal correlation degree. For each pair of sensors... and The correlation between them is calculated through spatiotemporal features and correlation indicators, and then a correlation matrix is ​​constructed.

[0101] Specifically, when calculating any element in the spatiotemporal correlation matrix, the first adjustment parameter is taken. Its value range is Second adjustment parameter Its value range is ;

[0102] Step 3: Dynamic Outlier Detection

[0103] In the preprocessing stage, based on the spatiotemporal correlation matrix Introducing dynamic thresholds Outlier detection is performed. This threshold can adaptively adjust based on changes in the correlation between different sensors, ensuring accurate detection of abnormal data in various scenarios. This is achieved by comparing standardized sensor data. With dynamic threshold If the difference between data points exceeds this threshold, the data is considered an outlier.

[0104] Specifically, when determining the dynamic threshold, the anomaly tolerance coefficient is included. The value is 0.5, and its range is [missing value]. ;

[0105] Step 4: Incremental Information Transmission and Data Repair

[0106] When outliers are detected, an incremental information transmission mechanism is used to repair the abnormal data. Based on the spatiotemporal correlation between sensors, information from adjacent sensors is used to weight and repair the abnormal data, ensuring that the repaired data can accurately reflect the spatiotemporal similarity between the sensors.

[0107] Step 5: Noise Reduction and Final Data Output

[0108] Noise is removed by weighted mean filtering, and the data is further smoothed to obtain the denoised detection data. The repaired data will serve as the basis for subsequent analysis.

[0109] ② Data Augmentation Based on Adaptive Temporal Multidimensional Information Augmentation and Intelligent Fusion Algorithm

[0110] Step 1: Nonlinear Temporal Modeling and Information Fusion

[0111] Preprocessed detection data Nonlinear time series modeling methods (such as nonlinear autoregressive models) are used to capture the dynamic characteristics in time series, and the detection data is predicted based on historical monitoring data and preprocessed detection data.

[0112] Specifically, in order to dynamically adjust the fusion weights of each sensor, an information entropy maximization mechanism is used to dynamically adjust the weights of each sensor during the multi-sensor data fusion process. This allows sensors with a large amount of information to play a greater role in data fusion.

[0113] Step 2: Dynamic adjustment of the gain function

[0114] Introducing an adaptive gain function By adjusting the data characteristics and gain of each sensor, the contribution of each sensor is matched with its information content, thereby improving the quality of data fusion.

[0115] Step 3: Data Fusion and Enhancement

[0116] With the adaptive gain function adjusted, data from multiple sensors are fused to obtain enhanced detection data. This data provides higher-quality input for subsequent fault detection, pattern recognition, and other processes.

[0117] In summary, a comprehensive performance detection data processing method based on sensor fusion has been developed.

[0118] Example 3 is the third embodiment of the present invention, which differs from the previous two embodiments in that:

[0119] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0120] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0121] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0122] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

Claims

1. A full performance detection data processing method based on sensor fusion, characterized in that: include, An adaptive spatiotemporal correlation denoising and repair algorithm is used to preprocess the multi-sensor detection data to obtain the preprocessed detection data. Anomaly detection is performed using a spatiotemporal correlation matrix to identify outliers in sensor data, and then repair and denoise them. The anomaly detection includes spatiotemporal feature decomposition representation based on sensor data. Establish a spatiotemporal correlation matrix Spatiotemporal correlation matrix Any element in the, i.e., the sensor With sensors At any moment The correlation is represented as: in, , and , They represent the first Spatial and temporal characteristics of each sensor It is the first adjustment parameter. It is the second adjustment parameter; Indicates sensor and sensors Between at a certain point in time Similarity on; Using the spatiotemporal correlation matrix Define dynamic threshold It detects outliers in sensor data. When the difference between the measured value of any sensor and the mean value of the corresponding spatiotemporal correlation neighborhood exceeds a dynamic threshold, the current data point is considered to be outlier. The repair and denoising include adjusting the standardized data. With dynamic threshold Comparison, If the value is 0, it is considered abnormal data. It is a sensor The mean; When an outlier is detected, an incremental information transmission mechanism is used for repair. This mechanism uses information from adjacent sensors to correct the outlier data. It is obtained by weighting the data differences with those of neighboring sensors; The existing weighted mean filter is used to remove noise, resulting in denoised detection data. ; The preprocessed detection data is intelligently enhanced using an adaptive temporal multidimensional information enhancement and intelligent fusion algorithm to obtain enhanced detection data.

2. The full-performance detection data processing method based on sensor fusion as described in claim 1, characterized in that: The preprocessing includes spatiotemporal feature decomposition and standardization of multi-source sensor data; Acquire data from different sensors, and combine data from each sensor. At the point of time Output measurement data, represented as a vector. ,in, It is the dimension of the data output by the sensor each time, and any element is used as... This indicates that it is the first The first sensor One data point, ; Spatiotemporal feature decomposition is performed on sensor data to extract spatial and temporal features and identify patterns of change in the data. The spatiotemporal feature decomposition is expressed as follows: in, It is the first Each sensor at a time point Spatiotemporal feature decomposition representation on, Indicates the first Each sensor at a time point Spatial features on, It is the first Each sensor at a time point The time characteristics on; Further analysis of arbitrary sensor data Normalization is performed to eliminate the influence of different sensors due to their different dimensions, resulting in standardized data. .

3. The full-performance detection data processing method based on sensor fusion as described in claim 2, characterized in that: The intelligent enhancement processing includes, based on the preprocessed detection data, using a nonlinear autoregressive model to establish a nonlinear time-series model to predict the detection data at the current moment. ; Calculate the entropy value of each sensor's data. The formula for measuring the diversity and uncertainty of information is: in, Indicates sensor At any moment The probability values ​​of the data distribution. It is the number of different states or values. It is a variable index; the higher the entropy value, the more information the sensor data contains.

4. The full-performance detection data processing method based on sensor fusion as described in claim 3, characterized in that: The intelligent enhancement processing also includes, based on entropy values, processing each sensor... At any moment Assign a dynamic weight The weight calculation formula is: in, For sensors At any moment The dynamic weights are defined as follows: n is the number of sensors, and i is the variable index.

5. The full-performance detection data processing method based on sensor fusion as described in claim 4, characterized in that: The intelligent enhancement processing also includes introducing an adaptive gain function to dynamically adjust the gain of each sensor data: in, It is a sensor The gain function of the sensor At any moment The gain value represents the adjustment factor for the sensor data; It is the first one after preprocessing Each sensor at time The mean, It is a sensor At any moment Standard deviation of data within a given time window It is a regularization factor.

6. The full-performance detection data processing method based on sensor fusion as described in claim 5, characterized in that: The intelligent enhancement processing also includes optimizing the fused and enhanced data based on an adaptive gain function. The enhanced detection data is specifically formulated as follows: in, It is enhanced detection data, which provides a foundation for subsequent data analysis.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the full-performance detection data processing method based on sensor fusion as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the full-performance detection data processing method based on sensor fusion as described in any one of claims 1 to 6.

Images