Multi-dimensional interactive computer big data fusion early warning method

Abstract

A multi-dimensional interactive computer big data fusion warning method belongs to the field of big data processing technology. The method includes: acquiring multi-source data and pre-processing the multi-source data to construct a fused data set. Calling an adaptive multi-level information filtering algorithm to segment the fused data set to output the filtered key warning information. Based on the key warning information, a multimodal interactive platform is constructed, and the multimodal interactive platform is used to output the control signal corresponding to each key warning information. In response to the control signal, warning feedback and control optimization are performed, and a feedback data matrix is ​​output. Among them, the multi-source data includes raw data from sensors of various dimensions, image acquisition devices and communication terminals, and the feedback data matrix is ​​composed of the response delay, instruction signal and correction parameters of each data interaction response. This method can ensure the rapid capture of key information in high-pressure scenarios and solve the limitations of a single interaction mode.

Description

Technical Field

[0001] The present invention belongs to the technical field of big data processing, and more specifically, relates to a multi-dimensional interactive computer big data fusion early warning method. Background Art

[0002] In extremely complex, multi-dimensional interactive display scenarios, warning information is often displayed centrally on a single platform. The fusion of data from different dimensions generates a vast amount of layered information. This "information stacking" can easily lead operators into cognitive bottlenecks in high-pressure work environments, making it difficult to accurately capture key anomaly indicators in a short period of time. This approach, also seen in air traffic control and astronomical observation centers, lacks effective dynamic filtering of hierarchical information within large-scale, high-density information displays.

[0003] Furthermore, existing technical solutions primarily rely on web-based and mobile terminal-based interaction. This fixed model struggles to meet the needs of skilled workers for real-time, convenient interaction in specialized work scenarios (e.g., in high-noise environments or when hands are limited during equipment operation). Therefore, timely feedback and evaluation of generated warning information are pressing challenges. Furthermore, existing multi-source information filtering techniques are relatively simple, making adaptive information filtering difficult to implement, resulting in a need for further improvement in the accuracy of capturing critical information. Summary of the Invention

[0004] In order to solve the deficiencies in the prior art, the purpose of the present invention is to solve the above-mentioned defects and further propose a multi-dimensional interactive computer big data fusion early warning method.

[0005] The present invention adopts the following technical solutions.

[0006] The first aspect of the present invention discloses a multi-dimensional interactive computer big data fusion early warning method, the method comprising:

[0007] Acquiring multi-source data and preprocessing the multi-source data to construct a fused data set;

[0008] Invoking an adaptive multi-level information filtering algorithm to segment the fused data set to output filtered key warning information;

[0009] Based on the key warning information, a multimodal interaction platform is constructed, wherein the multimodal interaction platform is used to output a control signal corresponding to each key warning information;

[0010] Performing early warning feedback and control optimization in response to the control signal, and outputting a feedback data matrix;

[0011] The multi-source data includes raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and the feedback data matrix is ​​composed of response delays, instruction signals, and correction parameters for each data interaction response.

[0012] Furthermore, the acquiring of multi-source data and preprocessing of the multi-source data to construct a fused data set includes:

[0013] Receive raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and integrate the raw data according to time sequence and data source to construct a raw data set;

[0014] Calibrate the original data set, and extract data columns corresponding to each device from the calibrated original data set;

[0015] Each data column is standardized based on its corresponding mean and standard deviation to construct a standardized feature matrix;

[0016] The data columns after the standardization processing of each device in the standardized feature matrix are fused and normalized to obtain the fused data set.

[0017] Furthermore, the adaptive multi-level information filtering algorithm is called to segment the fused data set to output filtered key warning information, including:

[0018] Performing data stratification and early warning indicator extraction on the fused data set to output a stratified feature matrix; elements in the stratified feature matrix represent early warning indicators at multiple different time points;

[0019] Based on the hierarchical feature matrix, each early warning indicator is weighted and summarized, and the weighted summary result is normalized to output an adaptive early warning score corresponding to the hierarchical feature matrix;

[0020] Based on the adaptive warning score, each hierarchical feature matrix in the fused data set is marked and screened to filter out the key warning information; the key warning information is a data column corresponding to the hierarchical feature matrix with an adaptive warning score not lower than a first threshold.

[0021] Furthermore, based on the key warning information, a multimodal interactive platform is constructed, and the multimodal interactive platform is used to output a control signal corresponding to each key warning information, including:

[0022] Converting the key warning information into a unified data format and performing feature decomposition on the key warning information to output a warning feature matrix;

[0023] Performing multimodal feedback parameter calculation on the warning feature matrix to obtain a multimodal feedback parameter matrix; the multimodal feedback parameters are visual feedback parameters, voice feedback parameters, gesture feedback parameters, and tactile feedback parameters output by the multimodal interaction platform based on the key warning information;

[0024] The multimodal feedback parameter matrix is ​​mapped to an allowable threshold range of each device through a mapping function, so as to convert the multimodal feedback parameter into the plurality of control signals.

[0025] Furthermore, the multimodal interaction platform is constructed based on the key warning information, and the multimodal interaction platform is used to output a control signal corresponding to each key warning information, and further includes:

[0026] Integrating the visual feedback parameters, voice feedback parameters, gesture feedback parameters, and tactile feedback parameters to obtain an interactive interface of the multimodal interaction platform;

[0027] Based on the control signal corresponding to each modal feedback parameter, a plurality of control modules are divided; the plurality of control modules include a vision module, a voice module, a gesture module and a tactile module;

[0028] Acquire operational feedback from the interactive interface in real time, and adaptively adjust the control signal and mapping function based on the operational feedback to establish a feedback mechanism;

[0029] The multiple control modules are written into the interactive interface, and combined with the feedback mechanism to construct the multimodal interactive platform.

[0030] Furthermore, the step of performing early warning feedback and control optimization in response to the control signal and outputting a feedback data matrix includes:

[0031] In response to the control signal, receiving warning feedback data corresponding to each key warning information from the interactive interface in real time; the warning feedback data includes the response delay, command signal and correction parameter corresponding to each key warning information;

[0032] Based on the warning feedback data, the warning trigger score corresponding to each key warning information in each data interaction response is calculated; the warning trigger score is used to evaluate the real-time warning response of each key warning information;

[0033] The early warning feedback data and multimodal feedback parameters corresponding to the critical early warning information with the early warning trigger score lower than the second threshold are adjusted to output the feedback data matrix.

[0034] Furthermore, the method further comprises:

[0035] Based on the warning trigger score, the key warning information is classified into risk levels according to a set threshold range to construct a warning data set consisting of multiple risk levels;

[0036] In response to the control signal, the risk level of the key warning information in the warning data set is output.

[0037] The second aspect of the present invention discloses a multi-dimensional interactive computer big data fusion early warning system, the system comprising:

[0038] A data preprocessing module, configured to acquire multi-source data and preprocess the multi-source data to construct a fused data set;

[0039] A data screening module is used to call an adaptive multi-level information filtering algorithm to segment the fused data set to output filtered key warning information;

[0040] An interactive platform construction module is used to construct a multimodal interactive platform based on the key warning information, and the multimodal interactive platform is used to output a control signal corresponding to each key warning information;

[0041] A feedback optimization module, configured to perform early warning feedback and control optimization in response to the control signal and output a feedback data matrix;

[0042] The multi-source data includes raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and the feedback data matrix is ​​composed of response delays, instruction signals, and correction parameters for each data interaction response.

[0043] A third aspect of the present invention discloses a terminal, comprising a processor and a storage medium, characterized in that:

[0044] The storage medium is used to store instructions;

[0045] The processor is configured to operate according to the instructions to execute the steps of the method of the first aspect.

[0046] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, wherein the program implements the steps of the method described in the first aspect when executed by a processor.

[0047] The beneficial effects of the present invention are that, compared with the prior art, the present invention has the following advantages:

[0048] (1) The present invention provides adaptive multi-level information filtering. To address the information stacking phenomenon, an adaptive filtering algorithm is used to perform hierarchical segmentation after data fusion. This not only improves the accuracy of capturing key information, but also ensures that operators can quickly capture key information in high-pressure scenarios, promotes the efficiency of the early warning model in generating early warning information, and thus improves the timeliness of feedback and response to the generated early warning information.

[0049] (2) The present invention provides a multimodal interactive presentation platform, which solves the limitation of a single interactive mode. By integrating multiple interactive modes, it can ensure real-time and convenient operation and response in various operating scenarios, and further improve the timeliness of feedback and response to the generated warning information in different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of the multi-dimensional interactive computer big data fusion early warning method provided by the present invention. DETAILED DESCRIPTION

[0051] The present application will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present application.

[0052] like Figure 1 As shown, in one embodiment, a multi-dimensional interactive computer big data fusion early warning method includes the following steps:

[0053] Step S110 , acquiring multi-source data and preprocessing the multi-source data to construct a fused data set.

[0054] Among them, multi-source data includes raw data from sensors of various dimensions, image acquisition devices and communication terminals.

[0055] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step S110 specifically includes the following steps:

[0056] Step S111 , receiving raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and integrating the raw data according to time sequence and data source to construct an original data set.

[0057] Step S112 , calibrating the original data set, and extracting the data columns corresponding to each device from the calibrated original data set.

[0058] Step S113 , performing standardization processing on each data column based on the mean and standard deviation corresponding to each data column to construct a standardized feature matrix.

[0059] Step S114 , fusing and normalizing the data columns of each device in the standardized feature matrix after the standardization process to obtain a fused data set.

[0060] In a specific embodiment, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention includes steps 1 to 4:

[0061] Step 1: Multi-source data collection and preprocessing.

[0062] Based on raw data from sensors, cameras (image acquisition devices), and communication terminals across various dimensions, we build a structured, standardized, and fused dataset. This addresses the cognitive overload caused by high-density information. By pre-cleaning, standardizing, and performing preliminary feature extraction on the data, we transform heterogeneous data into a unified format for subsequent layered filtering.

[0063] Specifically, the following steps are included:

[0064] Step 1.1: Data collection and raw data aggregation.

[0065] Specifically, raw data from sensors, cameras, and communication terminals in each dimension is collected and arranged uniformly by time and device type to form a raw data set R. The raw data set R is defined as a matrix with m rows and n columns, where m represents the number of sampling time points and n represents the number of different devices (sensors, cameras, and communication terminals). It can be expressed as: , where i = 1,2,...,m; j = 1,2,...,n. Represents the original data value collected by the jth device at the i-th time point.

[0066] Step 1.2, data synchronization and calibration.

[0067] Specifically, in order to eliminate the data deviation caused by time delay, calibration error and other issues of each device, the original data set R needs to be synchronized and calibrated. , using the following calibration formula: ,in, represents the data of the jth device at the i-th time point after calibration; is the offset of the jth device (determined according to the device's factory calibration data, usually a constant, ranging from 0 to 10); is the scaling factor of the jth device (used to normalize data of different units, recommended range is 0.1 to 10).

[0068] Step 1.3, feature extraction and data normalization.

[0069] Specifically, based on the calibrated data R', statistical features are extracted for each device (i.e., each column of data), and the mean and standard deviation corresponding to each data column are used for standardization to ensure comparability between data from different devices. For the jth device, the mean of the data column corresponding to the jth device is calculated. and standard deviation (The mean is the average value of the data at all time points of the device, and the standard deviation is the degree of data dispersion). Its expression is:

[0070]

[0071]

[0072] Where, Represents calibration data; is the average value of the jth device; is the standard deviation of the data of the jth device.

[0073] Standardization processing formula: / ,in, represents the normalized data of the jth device at the i-th time point. Normalization makes the data of each device follow a distribution with a mean of 0 and a variance of 1, which can effectively reduce the dimensionality of the data.

[0074] Step 1.4, data fusion and normalization.

[0075] Specifically, according to the standardized feature matrix Z, the standardized data of different devices are weightedly fused to generate a fused data set F that reflects the overall status.

[0076] Define the weighted fusion formula:

[0077]

[0078] in, is the fused data value at time point i; is the weight of the jth device (representing the reliability of the device data, determined based on historical accuracy and real-time stability, with a recommended value range of 0.1 to 1, with preference given to stable and accurate devices). is the data of the jth device at the i-th time point after standardization.

[0079] Step S120 , calling an adaptive multi-level information filtering algorithm to segment the fused data set to output filtered key warning information.

[0080] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step S120 specifically includes the following steps:

[0081] Step S121 , performing data stratification and early warning indicator extraction on the fused data set to output a stratified feature matrix; the elements in the stratified feature matrix represent early warning indicators at multiple different time points.

[0082] Step S122: Based on the hierarchical feature matrix, perform weighted aggregation on each early warning indicator, and normalize the weighted aggregation result to output an adaptive early warning score corresponding to the hierarchical feature matrix.

[0083] Step S123, marking and screening each hierarchical feature matrix in the fused data set based on the adaptive warning score to screen out key warning information; the key warning information is the data column corresponding to the hierarchical feature matrix with an adaptive warning score not lower than the first threshold.

[0084] In a specific embodiment, the multi-dimensional interactive computer big data fusion warning method provided by the present invention, step 2, adaptive multi-level information filtering algorithm processing, outputting key warning information after screening at each level, and targeting the information stacking phenomenon, using an adaptive filtering algorithm for hierarchical segmentation after data fusion to ensure that operators can quickly capture key information in high-pressure scenarios.

[0085] Specifically, the following steps are included:

[0086] Step 2.1: Data stratification and indicator feature extraction, outputting the stratified feature matrix P, where each element p_ik represents the kth warning indicator at the i-th time point.

[0087] Specifically, to reflect the role of each device data in early warning, the early warning indicator formula is defined:

[0088]

[0089] in, represents the kth indicator value at the i-th time point; is the value of the i-th row and k-th column in the fused dataset F; is the amplification factor of the kth indicator (determined by the sensitivity of the equipment, the recommended value range is 0.5 to 2); is the offset correction constant for the kth indicator (obtained from the equipment calibration data, with a recommended range of 0 to 5).

[0090] Finally, all Combined into hierarchical feature matrix .

[0091] Step 2.2, Adaptive early warning score calculation, output the early warning score vector S, where each element s_i represents the comprehensive early warning score at the i-th time point.

[0092] Specifically, to comprehensively consider the early warning effects of various equipment indicators, the weighted average method is used to calculate the early warning score for each time point i:

[0093]

[0094] Where, represents the comprehensive early warning score at the i-th time point; is the indicator obtained in sub-step 2.1; is the weight of the k-th indicator, reflecting the importance of the early warning response (the recommended value range is 0.1 - 1, determined according to the equipment historical data or expert experience).

[0095] The above formula realizes the weighted aggregation of each early warning indicator, and the normalized is convenient for formulating subsequent threshold judgments.

[0096] Step 2.3, Threshold judgment and early warning level division, output the early warning level marking vector L, where each element is the early warning risk classification at the i-th time point.

[0097] Specifically, set two key thresholds T1 and T2 to distinguish the levels of early warning risks, and require 0 < T1 < T2. The recommended value range of T1 is 0.7 - 1.0, and the recommended value range of T2 is 1.0 - 1.5. According to the comparison of s_i with T1 and T2, define the level division rules:

[0098] If < T1, then = "low risk";

[0099] If , then = "medium risk";

[0100] If , then = "high risk".

[0101] According to this rule, assign the corresponding risk level to each ​​​​​​​

[0103] Specifically, in order to highlight key information, a screening function is set for each risk level, the comprehensive score is weighted and modified, and the screening value is defined. :

[0104]

[0105] in, represents the revised score of the key warning information at time point i; Score the warning; is the hierarchical coefficient, defined as follows:

[0106] = 1.0, when = "High Risk";

[0107] = 0.5, when = "Medium Risk";

[0108] = 0.2, when = "Low risk".

[0109] The above formula makes the high-risk state and Keep close, and reduce at medium and low risk , thereby automatically screening out key warning information.

[0110] In this embodiment, in order to further improve the adaptive capability, a feedback adjustment formula is introduced to update the future scores. The expression is:

[0111]

[0112] in, is the early warning score at time point i after adjustment; is the original warning score; η is the learning rate (the recommended value range is 0.01-0.1, obtained by fitting historical response data); Score expectations set for high-risk objectives (e.g. Take between 1.2 and 1.5).

[0113] Through the above operations, all Composition set Q = { , , …, }, the score value in Q is the key warning information output after hierarchical screening.

[0114] Therefore, the present invention can gradually make the warning score more in line with the actual risk level through adaptive adjustment, and filter out the most representative warning information in the output Q.

[0115] Step S130: constructing a multimodal interaction platform based on the key warning information, where the multimodal interaction platform is used to output a control signal corresponding to each key warning information.

[0116] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step S130 specifically includes the following steps:

[0117] Step S131 : converting the key warning information into a unified data format and performing feature decomposition on the key warning information to output a warning feature matrix.

[0118] Step S132 , performing multimodal feedback parameter calculation on the warning feature matrix to obtain a multimodal feedback parameter matrix; the multimodal feedback parameters are visual feedback parameters, voice feedback parameters, gesture feedback parameters, and tactile feedback parameters output by the multimodal interaction platform based on the key warning information.

[0119] Step S133 : Mapping the multimodal feedback parameter matrix to the allowable threshold range of each device through a mapping function, so as to convert the multimodal feedback parameters into a plurality of control signals.

[0120] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step S130 specifically further includes the following steps:

[0121] Step S134 , integrating the visual feedback parameters, the voice feedback parameters, the gesture feedback parameters, and the tactile feedback parameters to obtain an interactive interface of the multimodal interactive platform.

[0122] Step S135 , dividing the control modules into multiple modules based on the control signals corresponding to the modal feedback parameters; the multiple control modules include a vision module, a voice module, a gesture module, and a tactile module.

[0123] Step S136 , obtaining operation feedback from the interactive interface in real time, and adaptively adjusting the control signal and the mapping function based on the operation feedback to build a feedback mechanism.

[0124] Step S137: Write multiple control modules into the interactive interface and combine them with the feedback mechanism to build a multimodal interactive platform.

[0125] In a specific embodiment, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step 3, constructs a multimodal interactive presentation platform. The platform is an interactive interface that integrates vision, voice, gesture and tactile feedback, which solves the limitations of a single interactive mode. By integrating multiple interactive methods, it ensures that technical workers can achieve real-time and convenient operation and response in various work scenarios.

[0126] Specifically, the following steps are included:

[0127] Step 3.1, data format conversion and feature decomposition, output warning feature matrix F = [ ], i = 1, 2,…, m; j = 1, 2,…, L.

[0128] in, represents the jth warning feature (such as risk level, response amplitude, time delay, etc.) extracted at the i-th time point, L is the total number of features, The expression is:

[0129]

[0130] in, The warning score for time point i; is the historical mean of the jth feature (obtained from historical data statistics, the recommended value range is 0 to 1); is the standard deviation of the jth feature (the recommended range is 0.1 to 0.5, calculated based on actual data).

[0131] Step 3.2, calculate the multimodal feedback parameters and output the multimodal feedback parameter matrix .

[0132] Among them, i = 1, 2, …, m; k = 1, 2, 3, 4 correspond to vision, speech, gesture and tactile feedback respectively.

[0133] Specifically, first calculate the comprehensive features (You can select weighted average of ):

[0134]

[0135] in, is the weight of the jth feature (the recommended range is 0.1 to 1, determined according to the feature importance).

[0136] After that, the expression is adopted for each modal feedback parameter:

[0137]

[0138] in, represents the feedback parameter of mode k at time point i; is the response magnification factor of mode k (for example, visual ,voice ,gesture , touch ); For basic response values ​​(e.g., visual ,voice ,gesture , touch ), determined by experience or preliminary testing.

[0139] Step 3.2 adaptively calculates the feedback parameters of each mode so that the warning information can be appropriately strengthened or attenuated in different interaction channels, preparing for further generation of control signals.

[0140] Step 3.3, mapping parameters and interaction signal generation. Interaction control signal matrix .

[0141] in, is the actual output signal parameter of mode k at time point i.

[0142] Specifically, first, a mapping function is used to ensure that the value is within the allowed range of the device. The specific formula is as follows:

[0143] (Visual Mapping)

[0144] (Voice Mapping)

[0145] (Gesture Mapping)

[0146] (Haptic Mapping)

[0147] in, 、 are the minimum and maximum thresholds for brightness adjustment of the visual device (e.g. = 0.1, = 1.0); 、 Set the volume threshold for the voice module (e.g. = 0.2, = 1.0); 、 The sensitivity threshold of the gesture module (recommended = 0.3, = 1.0); 、 is the tactile feedback intensity threshold (recommended = 0.0, = 1.0).

[0148] The function of step 3.3 is: the mapping function converts the feedback parameters into actual control signals to ensure that the output of each mode is within the acceptable range of the device, thereby achieving precise control and intuitive information presentation.

[0149] Step 3.4: Integrate the user interface construction and feedback closed loop to output an interactive interface that integrates vision, voice, gesture and tactile feedback.

[0150] Specifically, according to the control signal of each mode in E, each module is driven separately:

[0151] Vision module: The signal is used to dynamically adjust the brightness, color and animation effects of the warning information on the display screen;

[0152] Voice module: Based on Adjust the volume and speed of the warning voice broadcast;

[0153] Gesture Module: Utilize Adjust gesture recognition threshold and feedback sensitivity;

[0154] Tactile module: Adjust the vibration intensity of your wearable device.

[0155] Integrated control: The control signals of the above modules are unified through middleware to build a coordinated interactive interface to ensure the synchronous update of each modal information.

[0156] Feedback closed-loop mechanism: Real-time collection of operator feedback on the interface (such as response delay, confirmation operation, etc.), and adjustment of control signals or mapping functions through preset feedback formulas to improve interaction accuracy.

[0157] For example, to update the mapping parameters:

[0158]

[0159] Where η is the adjustment coefficient (0.01 to 0.1 is recommended), is the target response value, is the actual feedback response value.

[0160] In practical applications, at time point i = 100, the system obtains (Visual), (voice), (gesture), (tactile), the interface presents full screen brightness, high-volume voice broadcast, high-sensitivity gesture recognition, and medium vibration of wearable devices; if the feedback indicates that the voice broadcast is too loud, the system will appropriately reduce the volume according to the feedback formula , and then adjust , until the target response is reached.

[0161] In step 3.4, by converting the control signals of each modality into actual interactive interface outputs and establishing a feedback closed-loop mechanism, the interactive platform finally realizes a unified interactive interface integrating vision, voice, gesture and tactile feedback to meet the multimodal requirements in special operation scenarios.

[0162] Step S140 , performing early warning feedback and control optimization in response to the control signal, and outputting a feedback data matrix.

[0163] The feedback data matrix consists of the response delay, instruction signal and correction parameters for each data interaction response.

[0164] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step S140 specifically includes the following steps:

[0165] Step S141, in response to the control signal, receiving warning feedback data corresponding to each key warning information from the interactive interface in real time; the warning feedback data includes the response delay, instruction signal and correction parameter corresponding to each key warning information.

[0166] Step S142 , based on the warning feedback data, calculate the warning trigger score corresponding to each key warning information at each data interaction response; the warning trigger score is used to evaluate the real-time warning response of each key warning information.

[0167] Step S143: Adjust the warning feedback data and multimodal feedback parameters corresponding to the key warning information with a warning trigger score lower than the second threshold to output a feedback data matrix.

[0168] In some embodiments, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention further includes the following steps:

[0169] Step S210 , based on the warning trigger score, the key warning information is divided into risk levels according to the set threshold range to construct a warning data set consisting of multiple risk levels.

[0170] Step S220: In response to the control signal, output the risk level of the key warning information in the warning data set.

[0171] In a specific embodiment, the multi-dimensional interactive computer big data fusion early warning method provided by the present invention, step 4, immediate early warning feedback and closed-loop optimization, output user feedback data matrix .

[0172] in, Represents the i-th interaction response data, including response delay, confirmation signal, operation correction and other information.

[0173] Specifically, the operator's feedback data on the warning information is collected in real time on the interactive interface, including but not limited to the response time (Unit: seconds), confirmation mark (numeric type, 1 means confirmed, 0 means unconfirmed) and correction value (Indicates feedback correction of warning parameters, the recommended range is -1 to 1). Then combine the various feedbacks into a feedback vector ; All feedback data are arranged in sequence to form a matrix R, which is expressed as:

[0174]

[0175] in, is the response delay of the i-th user feedback; It is a confirmation sign, 1 means the operator confirms that the warning is valid; Adjust the value for feedback to reflect the user's correction suggestions for warning intensity or parameters.

[0176] Step 4.2: Real-time warning detection and trigger formula calculation, output of original warning trigger score vector .

[0177] Specifically, based on the feedback data, the warning trigger score is calculated for each interaction , set the scoring formula to:

[0178]

[0179] in, represents the warning trigger score of the i-th interaction; is the normalized response delay (the normalized value range is 0 to 1, and the normalization formula is: = ( − ) / ( − ),in 、 The shortest and longest response delays in history, respectively. = 0.1 s, = 1.0 s); Is the confirmation flag (confirmed value 1, so (1 − ) = 0, the value 0 is not confirmed, so 1); is the feedback correction value; w1, w2, w3 are the weights of each component respectively. The recommended values ​​are w1 = 0.4, w2 = 0.4, w3 = 0.2.

[0180] Step 4.2 comprehensively considers the response delay, confirmation status and feedback correction to evaluate the real-time effectiveness of the current warning response.

[0181] Step 4.3: Feedback closed-loop adjustment and dynamic parameter update to obtain the adjusted warning score vector Update values ​​with system parameters .

[0182] Specifically, in order to achieve closed-loop optimization, a feedback adjustment formula is used to dynamically update the early warning score, and the adjustment formula is set as follows:

[0183]

[0184] in, is the adjusted i-th warning score; Original score ; Target score value (set based on historical data and safety standards, recommended for high risk The range is 0.8 to 1.0, and can be set at 0.1 to 0.3 for low risk); η is the learning rate or adjustment coefficient (the recommended range is 0.01 to 0.1, obtained by fitting historical response data).

[0185] At the same time, each feedback parameter (for example, weights w1, w2, w3) is updated. By comparing the expected and actual effects, the weight update formula is defined:

[0186]

[0187] in, is the updated value of the kth weight; is the weight adjustment coefficient (0.01 to 0.1 is recommended); Rating the preset ideal feedback; Calculates the value for the current feedback.

[0188] The above formula helps the system adaptively optimize in subsequent calculations, gradually improving the accuracy of the early warning system. After such updates, the system will adjust parameters based on user feedback to ensure that the predicted scores are closer to the actual risks.

[0189] Step 4.4, the system generates and outputs real-time warning data.

[0190] Specifically, the system real-time warning data set ,in Indicates the final warning status at the i-th time point. Using the updated warning score, the warning information is finally divided according to the set threshold. Set threshold and ,suggestion = 0.3, = 0.7.

[0191] Defining alert states as follows:

[0192] like ,but = "low risk";

[0193] like ,but = "Medium Risk";

[0194] like ,but = "High risk".

[0195] All Integrate into real-time warning data sets .

[0196] The multi-dimensional interactive computer big data fusion early warning system provided by the present invention is described below. The multi-dimensional interactive computer big data fusion early warning system described below and the multi-dimensional interactive computer big data fusion early warning method described above can be referenced to each other.

[0197] In one embodiment, a multi-dimensional interactive computer big data fusion early warning system includes a data preprocessing module, a data screening module, an interactive platform construction module, and a feedback optimization module.

[0198] The data preprocessing module is used to obtain multi-source data and preprocess the multi-source data to construct a fused dataset.

[0199] The data screening module is used to call the adaptive multi-level information filtering algorithm to segment the fused data set to output the filtered key warning information.

[0200] The interactive platform construction module is used to build a multimodal interactive platform based on key warning information, and the multimodal interactive platform is used to output the control signal corresponding to each key warning information.

[0201] The feedback optimization module is used to perform early warning feedback and control optimization in response to the control signal and output a feedback data matrix.

[0202] Among them, multi-source data includes raw data from sensors of various dimensions, image acquisition devices and communication terminals, and the feedback data matrix is ​​composed of response delay, instruction signal and correction parameters during each data interaction response.

[0203] The present disclosure may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0204] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punched card or raised structure in a groove on which instructions are stored, and any suitable combination thereof. As used herein, a computer-readable storage medium is not to be construed as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse passing through a fiber optic cable), or an electrical signal transmitted through an electrical wire.

[0205] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions to be stored in the computer-readable storage medium in each computing / processing device.

[0206] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, the state information of the computer-readable program instructions is used to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), so that the electronic circuit can execute the computer-readable program instructions, thereby implementing various aspects of the present disclosure.

[0207] Various aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present disclosure. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer-readable program instructions.

[0208] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device is generated that implements the functions / actions specified in one or more blocks in the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium, where these instructions cause the computer, programmable data processing device, and / or other device to operate in a specific manner. Thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing various aspects of the functions / actions specified in one or more blocks in the flowchart and / or block diagram.

[0209] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operational steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more blocks in the flowchart and / or block diagram.

[0210] The flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and the part of the module, program segment or instruction contains one or more executable instructions for realizing the prescribed logical function. In some alternative implementations, the functions marked in the box can also occur in a sequence different from that 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 flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the prescribed function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0211] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A multi-dimensional interactive computer big data fusion early warning method, characterized in that: The method comprises: Acquiring multi-source data and preprocessing the multi-source data to construct a fused data set; Invoking an adaptive multi-level information filtering algorithm to segment the fused data set to output filtered key warning information; Based on the key warning information, a multimodal interaction platform is constructed, wherein the multimodal interaction platform is used to output a control signal corresponding to each key warning information; Performing early warning feedback and control optimization in response to the control signal, and outputting a feedback data matrix; The multi-source data includes raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and the feedback data matrix is ​​composed of the response delay, command signal, and correction parameters of each data interaction response; The calling of the adaptive multi-level information filtering algorithm to segment the fused data set to output filtered key warning information includes: Performing data stratification and early warning indicator extraction on the fused data set to output a stratified feature matrix; elements in the stratified feature matrix represent early warning indicators at multiple different time points; Based on the hierarchical feature matrix, each early warning indicator is weighted and summarized, and the weighted summary result is normalized to output an adaptive early warning score corresponding to the hierarchical feature matrix; Based on the adaptive warning score, each hierarchical feature matrix in the fused data set is marked and screened to filter out the key warning information; the key warning information is a data column corresponding to the hierarchical feature matrix with an adaptive warning score not lower than a first threshold.

2. The multi-dimensional interactive computer big data fusion early warning method according to claim 1 is characterized in that: The step of acquiring multi-source data and preprocessing the multi-source data to construct a fused data set includes: Receive raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and integrate the raw data according to time sequence and data source to construct a raw data set; Calibrate the original data set, and extract data columns corresponding to each device from the calibrated original data set; Each data column is standardized based on its corresponding mean and standard deviation to construct a standardized feature matrix; The data columns after the standardization processing of each device in the standardized feature matrix are fused and normalized to obtain the fused data set.

3. The multi-dimensional interactive computer big data fusion early warning method according to claim 2 is characterized in that: The multimodal interaction platform is constructed based on the key warning information, and the multimodal interaction platform is used to output a control signal corresponding to each key warning information, including: Converting the key warning information into a unified data format and performing feature decomposition on the key warning information to output a warning feature matrix; Performing multimodal feedback parameter calculation on the warning feature matrix to obtain a multimodal feedback parameter matrix; the multimodal feedback parameters are visual feedback parameters, voice feedback parameters, gesture feedback parameters, and tactile feedback parameters output by the multimodal interaction platform based on the key warning information; The multimodal feedback parameter matrix is ​​mapped to an allowable threshold range of each device through a mapping function, so as to convert the multimodal feedback parameter into the plurality of control signals.

4. The multi-dimensional interactive computer big data fusion early warning method according to claim 3 is characterized in that: The method further comprises: constructing a multimodal interactive platform based on the key warning information, wherein the multimodal interactive platform is configured to output a control signal corresponding to each key warning information; Integrating the visual feedback parameters, voice feedback parameters, gesture feedback parameters, and tactile feedback parameters to obtain an interactive interface of the multimodal interaction platform; Based on the control signal corresponding to each modal feedback parameter, a plurality of control modules are divided; the plurality of control modules include a vision module, a voice module, a gesture module and a tactile module; Acquire operational feedback from the interactive interface in real time, and adaptively adjust the control signal and mapping function based on the operational feedback to establish a feedback mechanism; The multiple control modules are written into the interactive interface, and combined with the feedback mechanism to construct the multimodal interactive platform.

5. The multi-dimensional interactive computer big data fusion early warning method according to claim 4 is characterized in that: The step of performing early warning feedback and control optimization in response to the control signal and outputting a feedback data matrix includes: In response to the control signal, receiving warning feedback data corresponding to each key warning information from the interactive interface in real time; the warning feedback data includes the response delay, instruction signal and correction parameter corresponding to each key warning information; Based on the warning feedback data, the warning trigger score corresponding to each key warning information in each data interaction response is calculated; the warning trigger score is used to evaluate the real-time warning response of each key warning information; The early warning feedback data and multimodal feedback parameters corresponding to the critical early warning information with the early warning trigger score lower than the second threshold are adjusted to output the feedback data matrix.

6. The multi-dimensional interactive computer big data fusion early warning method according to claim 5 is characterized in that: The method further comprises: Based on the warning trigger score, the key warning information is classified into risk levels according to a set threshold range to construct a warning data set consisting of multiple risk levels; In response to the control signal, the risk level of the key warning information in the warning data set is output.

7. A multi-dimensional interactive computer big data fusion early warning system, characterized by: The system comprises: A data preprocessing module, configured to acquire multi-source data and preprocess the multi-source data to construct a fused data set; The data screening module is used to call the adaptive multi-level information filtering algorithm to segment the fused data set to output the filtered key warning information; specifically, it includes: Performing data stratification and early warning indicator extraction on the fused data set to output a stratified feature matrix; elements in the stratified feature matrix represent early warning indicators at multiple different time points; Based on the hierarchical feature matrix, each early warning indicator is weighted and summarized, and the weighted summary result is normalized to output an adaptive early warning score corresponding to the hierarchical feature matrix; Based on the adaptive warning score, each hierarchical feature matrix in the fusion data set is marked and screened to screen out the key warning information; the key warning information is the data column corresponding to the hierarchical feature matrix with an adaptive warning score not lower than the first threshold. An interactive platform construction module is used to construct a multimodal interactive platform based on the key warning information, and the multimodal interactive platform is used to output a control signal corresponding to each key warning information; A feedback optimization module, configured to perform early warning feedback and control optimization in response to the control signal and output a feedback data matrix; The multi-source data includes raw data from sensors of various dimensions, image acquisition devices, and communication terminals, and the feedback data matrix is ​​composed of response delays, instruction signals, and correction parameters for each data interaction response.

8. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute 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 program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images