A multi-source data fusion method and system based on data analysis

Abstract

The application discloses a kind of multi-source data fusion method and system based on data analysis, it is related to electric digital data processing technical field.The multi-source data fusion method based on data analysis, it includes the following steps: fusion real-time analysis;Dynamic fusion adaptability analysis;Interaction fusion effectiveness analysis.The application carries out real-time analysis by real-time fusion data, judges whether sampling sensing frequency optimization is carried out, then according to the dynamic adaptability analysis of perception fusion data, judges whether perception fusion parameter optimization is carried out, finally carries out interaction effectiveness analysis, judges whether interaction fusion noise optimization is carried out, to realize the dynamic fusion reliability of corresponding fire information of fire-fighting facilities in substation in the process of internet of things communication is improved, the problem that the dynamic fusion reliability of corresponding fire information of fire-fighting facilities in substation in the process of internet of things communication is not high in prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, and in particular to a multi-source data fusion method and system based on data analysis. Background Technology

[0002] Existing technologies primarily collect environmental parameters and equipment status data through independent front-end devices such as fire sensors, temperature detectors, smoke alarms, fire hydrant pressure monitoring equipment, automatic sprinkler systems, and video surveillance cameras. After performing basic noise reduction and standardized format conversion on the raw data from each subsystem to achieve initial structuring, multi-source data is fused and calculated based on triggering mechanisms or shallow statistical analysis rules to generate local alarm signals. Specifically, fire hydrant pressure monitoring equipment detects external pressure information of fire hydrants, and automatic sprinkler systems include chemical agent adjustment equipment, water spray pumps, and nozzles. Subsequently, empirical secondary analysis of scattered alarm information is required to formulate fire response plans. Simultaneously, heterogeneous data unified modeling technology is used to construct a common data model framework, presenting the real-time spatial layout of substation fire protection facilities and integrating monitoring images of key areas, thereby improving the quality and efficiency of fire supervision and operational coordination capabilities.

[0003] For example, Chinese invention patent CN117828309B discloses a substation safety early warning method based on multi-source data fusion ranging, which includes: generating a virtual early warning space based on the danger distance threshold of the substation's high-voltage equipment, obtaining first monitoring data, activating a mobile ranging early warning module to perform tracking ranging if this value is less than or equal to the tracking distance threshold, obtaining a distance histogram time series diagram, combining the virtual early warning space with multi-source data fusion to generate a spatial contact probability, and issuing a safety early warning if this spatial contact probability is not less than the contact probability threshold.

[0004] For example, Chinese invention patent CN111209434B discloses a substation equipment inspection system and method based on multi-source heterogeneous data fusion, including: a multi-source heterogeneous data acquisition module, a video image data acquisition module and a sensor data acquisition module, and a multi-source heterogeneous data fusion module; wherein, the multi-source heterogeneous data acquisition module is used to convert and save the data acquired by the video image data acquisition module and the sensor data acquisition module, and the multi-source heterogeneous data fusion module contains a deep neural classification module and a classification result fusion module.

[0005] The above-mentioned technology has at least the following technical problems:

[0006] In existing technologies, during initial fire monitoring, when a sudden increase in combustible vapor concentration causes a rapid change in the temperature rise rate, the temperature sensors covering the main transformer area, lacking real-time sensing technology, suffer from insufficient thermal inertia compensation, resulting in measured values ​​lagging behind actual values. Secondly, fire-fighting gas diffusion requires time, and the combustion products (smoke) include smoke particles and harmful gases; the delay caused by gas diffusion during the mixed gas reaction process is not compensated for. Furthermore, the water spray equipment located near the main transformer area maintains conventional atomization parameters and fails to switch to high-speed flow in time, making it impossible to adjust the water mist characteristics within the atomization equipment in real time according to the characteristics of the combustion product smoke. However, the system fails to integrate real-time changes in existing environmental data and lacks the ability to compensate for the control window of fire spread caused by response delays in the early stages of a fire. In high-voltage substation scenarios, electromagnetic pulses caused by high-current switching actions cause a sharp drop in the signal-to-noise ratio of wireless transmission channels. Rapid changes in current generate strong magnetic field changes, which in turn lead to changes in the electric field, thus forming electromagnetic pulses. These pulses are radiated onto communication lines, causing transient interference, which in turn reduces data coverage and causes network jitter. This results in a problem of low reliability in the dynamic integration of fire protection information by fire protection facilities in substations during IoT communication. Summary of the Invention

[0007] This invention provides a multi-source data fusion method and system based on data analysis, which solves the problem of low reliability of dynamic fusion of fire protection information corresponding to fire protection facilities in substations during IoT communication in the prior art, and improves the reliability of dynamic fusion of fire protection information corresponding to fire protection facilities in substations during IoT communication.

[0008] This invention provides a multi-source data fusion method based on data analysis, comprising the following steps: A fire monitoring and management platform receives fire information data sent by an IoT terminal; based on the acquired real-time fusion data, it performs real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform to determine whether to optimize the sampling sensor frequency. Sampling sensor frequency optimization means improving the real-time synchronization of fire information data fusion by adjusting the checkpoint interval and node density. If the synchronization fusion is deemed successful, a dynamic adaptive analysis is performed on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether to optimize the sensing fusion parameters. Sensing fusion parameter optimization means improving the effectiveness of sensing fusion of fire information data by compensating for the window period caused by the sensing fusion response delay. If the sensing fusion is deemed successful, an interaction effectiveness analysis is performed on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether to optimize the interaction fusion noise. Interaction fusion noise optimization means improving the effectiveness of interaction fusion of fire information data by reducing noise interference during the interaction fusion process.

[0009] This invention provides a multi-source data fusion system based on data analysis, comprising: a real-time fusion analysis module, a dynamic fusion adaptability analysis module, and an interactive fusion effectiveness analysis module. The real-time fusion analysis module is used by the fire monitoring and management platform to receive fire information data sent by IoT terminals and to perform real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform based on the acquired real-time fusion data to determine whether sampling sensor frequency optimization is required. The dynamic fusion adaptability analysis module is used to perform dynamic adaptive analysis on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether sensing fusion parameter optimization is required. The interactive fusion effectiveness analysis module is used to perform interactive effectiveness analysis on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether interactive fusion noise optimization is required.

[0010] One or more technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0011] 1. By analyzing real-time fused data, sensing fused data, and the stabilization time of mixed gas, sampling sensor frequency optimization, sensing fusion parameter optimization, and interactive fusion noise optimization are triggered respectively. This allows fire sensors to detect abnormal information and transmit it to relevant fire equipment in real time. The real-time fused data is compared and corrected with the database to generate accurate scores. By coupling and quantifying real-time performance, the accuracy of delay assessment is improved. Sensing fusion parameters are optimized using sensing fusion data to dynamically compensate for delays caused by response. The stabilization time of mixed gas is used to evaluate the effectiveness of interaction. The substation fire information reception has near-zero latency, equipment operation has near-zero error, and safety is better guaranteed.

[0012] 2. By comparing real-time fused data with preset data and correcting it with correction values, a real-time fused data score is generated. Coupled processing generates an indicator of the impact of data fusion real-time performance, accurately quantifying the degree of impact of data fusion real-time performance. This processing improves the accuracy of data fusion delay acquisition. The correction value eliminates inherent biases, making the acquired score closer to the true level. Coupled processing of multi-dimensional characteristics constructs a scientific quantitative indicator. The indicator itself can reflect the degree of time loss caused by reception during data fusion, providing a decision-making basis for optimizing sampling sensor frequency and ensuring the accuracy of delay assessment for real-time data fusion.

[0013] 3. Improve the control accuracy of fire-fighting equipment by constructing a dynamic perception fusion parameter optimization system; use the integral method to process and evaluate the time synchronization error between the sensor and the fire-fighting execution equipment, and couple to generate perception fusion impact indicators to accurately quantify the time sequence matching degree of multi-source perception data streams; this integral processing improves the accuracy of synchronization error assessment, smooths random fluctuations, and makes the indicators closer to the real synchronization state; this indicator can reflect the degree of response delay caused by synchronization error before the flame growth inflection point, providing a decision-making basis for extinguishing initial smoldering fires, ensuring the rapid formation of a chemical concentration advantage field, avoiding the risk of reignition of oil-immersed equipment, and improving the overall reliability of fire-fighting response. Attached Figure Description

[0014] Figure 1 A flowchart of a multi-source data fusion method based on data analysis provided in an embodiment of the present invention;

[0015] Figure 2 A flowchart of the fusion real-time analysis and determination process provided in this embodiment of the invention;

[0016] Figure 3 A flowchart of dynamic fusion adaptability analysis and determination provided in an embodiment of the present invention;

[0017] Figure 4 A flowchart for the interaction fusion effectiveness analysis and determination provided in this embodiment of the invention;

[0018] Figure 5 This is a schematic diagram of the structure of a multi-source data fusion system based on data analysis provided in an embodiment of the present invention. Detailed Implementation

[0019] This invention provides a multi-source data fusion method and system based on data analysis, which solves the problem of low reliability of dynamic fusion of fire information corresponding to fire protection facilities in substations during IoT communication in the prior art. The method involves a fire monitoring and management platform receiving fire information data sent by IoT terminals, and performing real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform based on the acquired real-time fusion data to determine whether sampling sensor frequency optimization is needed. If the synchronous fusion is deemed successful, dynamic adaptive analysis is performed on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether sensing fusion parameter optimization is needed. If the sensing fusion is deemed successful, interactive effectiveness analysis is performed on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether interactive fusion noise optimization is needed. This improves the reliability of dynamic fusion of fire information corresponding to fire protection facilities in substations during IoT communication.

[0020] The technical solution in this invention aims to address the problem of low reliability in the dynamic fusion of fire protection information corresponding to fire protection facilities in substations during IoT communication. The overall approach is as follows:

[0021] The necessity of optimizing the sampling sensing frequency is determined based on the real-time fusion data analysis results. The dynamic adaptability is evaluated based on the steel fusion data to determine the optimization requirements of the sensing fusion parameters. The effectiveness of interactive fusion is analyzed by the stable duration of the mixed gas, and the implementation conditions for interactive fusion noise optimization are determined. This achieves the effect of improving the reliability of dynamic fusion of fire protection information in the IoT communication process of fire protection facilities in substations.

[0022] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0023] like Figure 1 The diagram shows a flowchart of a multi-source data fusion method based on data analysis provided in an embodiment of the present invention. This method includes the following steps: A fire monitoring and management platform receives fire information data sent by an IoT terminal. Based on the acquired real-time fusion data, it performs real-time analysis on the fusion process between the automatic fire alarm device and the fire monitoring and management platform to determine whether to optimize the sampling sensor frequency. Sampling sensor frequency optimization means improving the real-time synchronization of fire information data fusion by adjusting the checkpoint interval and node density. If the synchronization fusion is deemed successful, a dynamic adaptive analysis is performed on the fusion process between the transformer water spray device and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether to optimize the sensing fusion parameters. Sensing fusion parameter optimization means improving the effectiveness of sensing fusion of fire information data by compensating for the window period caused by the sensing fusion response delay. If the sensing fusion is deemed successful, an interaction effectiveness analysis is performed on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether to optimize the interaction fusion noise. Interaction fusion noise optimization means improving the effectiveness of interaction fusion of fire information data by reducing noise interference during the interaction fusion process.

[0024] In this embodiment, key components such as fire protection facilities and fire alarm systems are connected through IoT sensing and communication technologies. The real-time operation status of key components of substation fire protection equipment, including automatic fire alarm equipment, transformer water spray equipment, fire gas extinguishing equipment, fire hydrant water pressure monitoring, and fire water tank water level monitoring, is displayed. This reduces transmission latency, improves sampling coverage, ensures efficient synchronization of terminal information with the platform, and enhances the real-time decision-making capability for fire alarm response. In particular, the real-time performance, adaptability, and interactive effectiveness of fire data fusion are optimized in stages to improve the efficiency of smart fire protection.

[0025] Furthermore, based on the acquired real-time fusion data, a real-time performance analysis is performed on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform. Specific steps include: comparing the acquired real-time fusion data with preset real-time fusion data in the database, and simultaneously performing correction processing using real-time fusion data correction values ​​to obtain a real-time fusion data score; coupling the data coverage rate score, network jitter rate score, and electromagnetic interference intensity score to obtain the data fusion real-time performance impact index; the real-time fusion data includes data coverage rate, network jitter rate, and electromagnetic interference intensity. The data coverage rate is recorded using a power quality analyzer. The network jitter rate was obtained through one-way delay jitter calculation, and the electromagnetic interference intensity was measured using an oscilloscope spectrum analyzer. The data fusion real-time performance impact index represents the quantitative data on the impact of real-time fused data on the real-time fusion performance between the automatic fire alarm equipment and the fire monitoring and management platform. The preset real-time fused data includes preset data coverage rate, preset network jitter rate, and preset electromagnetic interference intensity. The real-time fused data correction values ​​include data coverage rate correction values, network jitter rate correction values, and electromagnetic interference intensity correction values. The real-time fused data score includes data coverage rate score, network jitter rate score, and electromagnetic interference intensity score.

[0026] The specific expression for the data coverage score Q1 is as follows: In the formula, Q1 represents the data coverage rate score at the end of the integration period between the automatic fire alarm equipment and the fire monitoring and management platform, C1 represents the data coverage rate correction value, and Q C1 Q represents the data coverage rate at the end of the real-time analysis period of the fusion process. C0 This indicates the preset data coverage rate.

[0027] The specific expression for the network jitter score Q2 is: In the formula, Q2 represents the network jitter score at the end of the fusion period between the automatic fire alarm equipment and the fire monitoring and management platform, C2 represents the network jitter correction value, and Q... C2 Q represents the network jitter rate at the end of the real-time analysis period of the fusion process. N0 This indicates the preset network jitter rate.

[0028] The specific expression for the electromagnetic interference intensity fraction Q3 is: In the formula, Q3 represents the electromagnetic interference intensity score at the end of the fusion period between the automatic fire alarm equipment and the fire monitoring and management platform, C3 represents the electromagnetic interference intensity correction value, and Q C3 Q represents the electromagnetic interference intensity at the end of the real-time analysis period of the fusion process. E0 This indicates the preset electromagnetic interference intensity.

[0029] Data fusion real-time impact index R DF The specific expression is: In the formula, R DF This indicates the real-time impact index of data fusion at the end of the fusion period between the automatic fire alarm equipment and the fire monitoring and management platform.

[0030] In this embodiment, the preset real-time fusion data is represented by the summation and averaging of the real-time fusion data corresponding to the end of the historical fusion period between the automatic fire alarm equipment and the fire monitoring management platform. The real-time fusion data is obtained from the real-time fusion data corresponding to the end of the fusion period between the automatic fire alarm equipment and the fire monitoring management platform. The data coverage rate correction value, network jitter rate correction value, and electromagnetic interference intensity correction value are preset values ​​in the database used to measure the degree of influence of data coverage rate, network jitter rate, and electromagnetic interference intensity on the acquisition of the data fusion real-time impact index. The database stores the corresponding correction values ​​for each parameter, and there is a preset mapping relationship between them, which can be many-to-one or one-to-one. In practical applications, inputting the real-time data coverage rate, network jitter rate, and electromagnetic interference intensity can accurately obtain the corresponding correction values, providing a quantitative basis for evaluating the impact of the data fusion real-time impact index on the fusion degree, and helping to accurately calculate the data fusion real-time impact index. The values ​​of these three correction values ​​are all in the range of 0-1, and their sum equals 1.

[0031] It is important to note that the impact of data fusion real-time performance metrics increases with the increase of data coverage rate, network jitter rate, and electromagnetic interference intensity. The decrease in data coverage rate means that when multiple nodes transmit data concurrently in the same area, channel contention intensifies, increasing the volatility of network transmission latency, i.e., the network jitter rate rises. If high electromagnetic interference exists in the environment at this time, the superimposed interference signal will increase the bit error rate of wireless communication, further increasing the probability of data packet retransmission, forcing the reduction of sampling frequency or coverage area to maintain communication stability. While low-frequency sampling can reduce network jitter rate, it will lead to an increase in data coverage rate, weakening the ability to monitor the status of critical areas.

[0032] By considering the above-mentioned mutual influence mechanisms, we can gain a more comprehensive understanding of the relationship between the real-time impact indicators of data fusion and various variables. These relationships are crucial for real-time assessment during the acquisition of real-time impact indicators of data fusion. By optimizing the data coverage rate, network jitter rate, and electromagnetic interference intensity, we have improved the integration between automatic fire alarm equipment and fire monitoring and management platform, effectively solving the problem of low reliability of dynamic integration of fire information in fire protection facilities in substations during IoT communication.

[0033] like Figure 2The diagram shows a flowchart of the real-time performance analysis and judgment process provided in this embodiment of the invention. It determines whether the acquired data fusion real-time performance impact index is greater than a preset value. If it is, the optimization process is initiated. The first step involves checking and adjusting the checkpoint interval, then re-acquiring the index value and determining whether it is not greater than a preset index in the database. If it is not greater, dynamic adaptive analysis is performed; if it is greater, sampling sensor frequency optimization is performed, and the index value is acquired again. When the final acquired index value is not greater than the preset value, the process ends; otherwise, the adjustment process continues cyclically. The entire optimization cycle must be completed within a specified number of iterations; if this limit is exceeded, a sampling sensor frequency warning is triggered.

[0034] Further understanding is needed regarding the determination of whether to perform sampling sensor frequency optimization. The specific steps are as follows: compare the acquired data fusion real-time impact index with the preset data fusion real-time impact index in the database. If the acquired data fusion real-time impact index is greater than the preset data fusion real-time impact index in the database, it is recorded as data synchronization failure and sampling sensor frequency optimization is performed. If the acquired data fusion real-time impact index is not greater than the preset data fusion real-time impact index in the database, it is recorded as data synchronization success and dynamic adaptive analysis is performed.

[0035] The specific steps for optimizing the sampling sensor frequency are as follows: First, optimize the checkpoint interval: Obtain the deviation of the real-time impact index of data fusion and the deviation of the alarm response time, respectively, and perform proportional processing with the preset deviation of the real-time impact index of data fusion and the deviation of the alarm response time in the database. Sum and average the results after proportional processing to obtain the reserve value of the checkpoint interval, which is used to prompt the scheduler to adjust the checkpoint interval based on the obtained reserve value of the checkpoint interval to suppress the occurrence of glitches in the signal transmission band. After optimizing the sampling sensor frequency, determine whether the deviation of the real-time impact index of data fusion that is re-obtained is not greater than the preset deviation of the real-time impact index of data fusion in the database. If so, complete the first step and perform dynamic adaptive analysis; otherwise, proceed to the second step.

[0036] The second step is to optimize node density: Based on the newly acquired node density reserve value, the topology self-organizing engine adjusts the node density to reduce the probability of phase distortion. The node density reserve value represents the deviation of the real-time impact index of data fusion and the deviation of the alarm response time after the sampling sensor frequency optimization is re-acquired. These deviations are proportionally processed and then summed and averaged with the preset deviations of the real-time impact index of data fusion and the alarm response time in the database. After the sampling sensor frequency optimization, if the newly acquired real-time impact index of data fusion is not greater than the preset real-time impact index of data fusion in the database, the second step is completed and dynamic adaptive analysis is performed; otherwise, the number of sampling sensor frequency optimization attempts is judged. The specific steps for judging the number of sampling sensor frequency optimization attempts are as follows: if the number of sampling sensor frequency optimization attempts is within the specified number of attempts, the first step is returned; otherwise, a sampling sensor frequency warning is issued.

[0037] In this embodiment, an adaptive adjustment algorithm is used to adjust the checkpoint interval parameter in real time based on the acquired checkpoint interval to achieve smooth signal transmission. Simultaneously, historical checkpoint interval pre-set values ​​are used as sample data and input into the autoregressive model of the historical load data time-series prediction model. The checkpoint interval-node density time-series prediction model is trained based on the adaptive adjustment algorithm. The currently acquired checkpoint interval pre-set values ​​and node density pre-set values ​​are input into the checkpoint interval-node density time-series prediction model, and the corresponding adjusted checkpoint interval values ​​and node density values ​​are output. The deviation of the data fusion real-time impact index represents the difference between the acquired data fusion real-time impact index and the preset data fusion real-time impact index in the database. The alarm response time deviation represents the difference between the acquired alarm response time and the preset alarm response time in the database.

[0038] This example establishes a hierarchical and progressive sampling sensor frequency optimization mechanism to achieve intelligent optimization of the fire data synchronization and fusion process, suppress high-frequency glitches in the transmission signal, and make the real-time data stream exhibit smoother transmission characteristics. It maintains data integrity by reducing the probability of phase distortion caused by multipath effects. At the same time, it establishes a closed-loop convergence mechanism for data fusion error within a preset number of optimizations, forming a dynamic balance between fire alarm response delay and data packet loss rate, thereby achieving an average reduction in fusion delay, a decrease in the actual occurrence rate of phase distortion, and an improvement in concurrent transmission stability.

[0039] Furthermore, based on the acquired sensing fusion data, a dynamic adaptive analysis is conducted on the fusion process between the transformer water spray equipment and the fire monitoring and management platform. Specific steps include: performing integral processing on the acquired sensing fusion data within a defined sensing fusion time interval to obtain the sensing fusion data area; simultaneously, coupling processing is performed on the sensing fusion data area to obtain sensing fusion impact indicators. The sensing fusion time interval represents the unit time period used to analyze sensing fusion efficiency (corresponding to a rapidly changing fire environment; this unit time period is typically set to a few seconds). The integral processing represents the quantification of the cumulative changes in sensing fusion data within the sensing fusion time interval. The sensing fusion data includes changes in temperature rise rate, changes in combustion product smoke concentration, and changes in smoke diffusion speed. The larger these three changes are, the more complex the fire environment, which will exacerbate data interference and conflicts, thereby reducing the accuracy and real-time performance of sensing fusion between the equipment and the platform. The sensing fusion data area includes the area of ​​changes in temperature rise rate, the area of ​​changes in combustion product smoke concentration, and the area of ​​changes in smoke diffusion speed.

[0040] Among them, the temperature rise rate change reflects the temperature change during combustion within the sensing fusion time interval; the combustion product smoke concentration change reflects the smoke concentration change during combustion within the sensing fusion time interval; the smoke diffusion rate change reflects the smoke diffusion during combustion within the sensing fusion time interval; the area of ​​the temperature rise rate change represents the area under the integral curve corresponding to the temperature rise rate change over time within the sensing fusion time interval, used to quantify the cumulative effect of temperature change; the area of ​​the combustion product smoke concentration change represents the area under the integral curve corresponding to the smoke concentration change over time within the sensing fusion time interval, used to quantify the cumulative effect of smoke generation; the area of ​​the smoke diffusion rate change represents the area under the integral curve corresponding to the smoke diffusion rate change over time within the sensing fusion time interval, used to quantify the cumulative effect of smoke diffusion; and the sensing fusion impact index represents the quantitative data on the degree of dynamic adaptive impact of sensing fusion data on the fusion process between the transformer water spray equipment and the fire monitoring and management platform.

[0041] In this embodiment, by constructing a perception fusion effectiveness evaluation method based on multi-parameter spatiotemporal integral modeling, the dynamic collaborative efficiency of the fire monitoring platform and the fire extinguishing device is improved: by using multi-dimensional data integral calculation of temperature rise rate, smoke concentration and diffusion speed, the fire characteristic parameters within the second-level time window are transformed into quantifiable area indicators, which can accurately characterize the linkage and coupling strength of multi-sensor data during the fire development process. Then, based on the area ratio weight of each parameter, the response parameter matrix of the water spray equipment is optimized, thereby improving the accuracy of early fire extinguishing agent delivery and reducing the false spray rate, and blocking the chain reaction of transformer oil conservator deflagration.

[0042] like Figure 3The diagram shows a flowchart of the dynamic fusion adaptive analysis and judgment process provided in this embodiment of the invention. It determines whether the acquired perception fusion impact index is greater than a preset value. If it is, optimization is initiated: first, data alignment is corrected to adjust the perception fusion response delay; then, the index value is re-evaluated. If the corrected index does not meet the preset value requirement, the process enters the traffic adjustment stage, updating parameters through the perception fusion delay compensation mechanism and detecting the index again. If the index value is not greater than the preset value, the optimization instruction is completed. If the index value is greater than the preset value, it is first determined whether it is within the specified number of optimization attempts. If it is within the specified number of optimization attempts, the process is restarted; if it is not within the specified number of optimization attempts, a perception fusion warning is issued.

[0043] Further, it is necessary to understand that the specific steps for determining whether to optimize the perception fusion parameters are as follows: compare the obtained perception fusion impact index with the preset perception fusion impact index in the database: if the obtained perception fusion impact index is not greater than the preset perception fusion impact index in the database, then the perception fusion is considered qualified and interaction effectiveness analysis is performed; if the obtained perception fusion impact index is greater than the preset perception fusion impact index in the database, then the perception fusion is considered unqualified and perception fusion parameter optimization is performed.

[0044] The specific steps for optimizing the perception fusion parameters are as follows: First, data alignment correction: Obtain the deviation of the perception fusion impact index and the deviation of the inference delay time, respectively, and proportionally process them with the preset deviations of the perception fusion impact index and the inference delay time in the database. Sum and average the results after proportional processing to obtain the data alignment correction value, which is used to indicate the time loss caused by the injection controller due to the perception fusion response delay. The inference delay time represents the delay time generated by the smoke sensor in smoke recognition during combustion. After data alignment correction, determine whether the newly acquired perception fusion impact index deviation is not greater than the preset perception fusion impact index deviation in the database. If so, complete the first step and perform interaction validity analysis; otherwise, it indicates response delay overload. Then proceed to the second step; the second step is to perform perception fusion delay compensation: based on the reacquired agent injection flow rate adjustment value, increase the agent injection flow rate to improve the control effectiveness of the fire in the early stage of the fire. The agent injection flow rate adjustment value represents the harmonic average result of the reacquired perception fusion impact index deviation and the data alignment deviation after data alignment correction; after the perception fusion parameters are optimized, if the reacquired perception fusion impact index is not greater than the preset perception fusion impact index in the database, then the second step is completed and interactive effectiveness analysis is performed; otherwise, the number of perception fusion parameter optimizations is judged: the specific steps for judging the number of perception fusion parameter optimizations are as follows: if the number of perception fusion parameter optimizations is within the specified number of perception fusion parameter optimizations, then return to the first step; otherwise, a perception fusion warning is issued.

[0045] In this embodiment, the perception fusion impact index deviation represents the difference between the acquired perception fusion impact index and the preset perception fusion impact index; the data alignment deviation represents the difference between the acquired data alignment and the preset data alignment in the database; and the inference delay duration deviation represents the difference between the acquired inference delay duration and the preset inference delay duration in the database. The data alignment is adjusted in a timely manner based on the acquired current data alignment correction value using a timestamp synchronization algorithm to compensate for the time delay in the fusion process. Simultaneously, historical data alignment correction values ​​and flow values ​​are used as sample data and input into the multiple linear regression model in the linear regression model. A data alignment-flow adjustment linear regression model is obtained by training based on the timestamp synchronization algorithm. The acquired data alignment correction values ​​and flow values ​​are then input into the data alignment-flow adjustment linear regression model, ultimately outputting the corresponding data alignment correction adjustment values ​​and flow adjustment values.

[0046] This example improves the control accuracy of fire-fighting equipment by constructing an optimization system for dynamic sensing and fusion parameters, dynamically corrects the time synchronization error between sensor data and fire-fighting execution equipment, improves the timing matching degree of multi-source sensing data streams, ensures the rapid formation of a chemical concentration advantage field before the flame growth inflection point, improves the timeliness of extinguishing initial smoldering fires, avoids the risk of reignition in oil-immersed equipment, and improves the overall reliability of fire-fighting response.

[0047] like Figure 4 The diagram shows the flowchart for the interaction fusion effectiveness analysis and judgment provided in this embodiment of the invention. It determines whether the obtained stable time of the mixed gas is greater than a preset value. If so, optimization is initiated: first, the interaction fusion delay is corrected, and then the time is re-detected. If the re-obtained index is not greater than the preset value in the database, the optimization is completed; otherwise, the fusion noise suppression value is increased and the time is re-evaluated. The index is re-obtained. If the index is not greater than the preset value, interaction effectiveness analysis is performed. If it is not within the specified number of optimization attempts, an interaction fusion warning is sent.

[0048] Further understanding is needed regarding the specific steps for conducting an interactive effectiveness analysis of the fusion process between fire extinguishing equipment and the fire monitoring and management platform based on the obtained stable gas mixture duration, in order to determine whether interactive fusion noise optimization should be performed. These steps involve comparing the obtained stable gas mixture duration with the preset stable gas mixture duration in the database. If the obtained stable gas mixture duration is longer than the preset stable gas mixture duration in the database, it is considered unstable and interactive fusion parameter optimization is performed. If the obtained stable gas mixture duration is not greater than the preset stable gas mixture duration in the database, it is considered stable and the interactive effectiveness analysis is completed.

[0049] The specific steps for optimizing the interactive fusion parameters are as follows: First, perform interactive fusion delay optimization: obtain the mixed gas stabilization time deviation and the fire gas release rate deviation, respectively, and perform proportional processing with the preset mixed gas stabilization time deviation and fire gas release rate deviation in the database. The results of the proportional processing are summed and averaged to obtain the interactive fusion delay correction value, which is used to correct the delay generated by the gas diffusion device, promote faster and more uniform distribution of fire gas, and improve the response speed of the gas diffusion device. After the interactive fusion delay correction, it is determined whether the newly acquired gas mixture stabilization time is not greater than the preset gas mixture stabilization time in the database. If so, the first step is completed and interactive validity analysis is performed; otherwise, the second step is performed. The second step is to optimize the fusion noise: the deviation of the gas mixture stabilization time after the interactive fusion delay correction is mapped in the database to obtain the fusion noise enhancement value, which is used to improve the noise amplitude suppression effect of the corresponding noise covariance matrix in the Kalman filter algorithm. After the interactive fusion parameter optimization, if the newly acquired gas mixture stabilization time is not greater than the preset gas mixture stabilization time in the database, the second step is completed and interactive validity analysis is performed; otherwise, the number of interactive fusion parameter optimizations is judged. The specific steps for judging the number of interactive fusion parameter optimizations are as follows: if the number of interactive fusion parameter optimizations is within the specified number of interactive fusion parameter optimizations, the first step is returned; otherwise, an interactive fusion warning is issued.

[0050] Among them, the gas stabilization time represents the time required for the gas mixture to reach stability after the reaction between the fire extinguishing gas and the combustion products; the gas stabilization time deviation represents the difference between the obtained gas stabilization time and the preset gas stabilization time; the gas stabilization time is obtained by monitoring through a built-in timer; the preset gas stabilization time is represented by the sum and average of the historical gas stabilization times of the fusion process between historical fire extinguishing equipment and fire monitoring and management platform; and the fire extinguishing gas release rate deviation represents the difference between the obtained fire extinguishing gas release rate and the preset fire extinguishing gas release rate.

[0051] In this embodiment, a parameter adaptive algorithm is used to compensate for the current interaction fusion delay in real time, reducing the window period caused by the need for gas diffusion. At the same time, the fusion noise enhancement value is used as sample data and input into the Bayesian network model. The interaction fusion-noise suppression Bayesian network model is trained based on the parameter adaptive algorithm. The obtained interaction fusion delay correction value and fusion noise enhancement value are input into the interaction fusion-noise suppression Bayesian network model, and finally the corresponding interaction fusion delay correction adjustment value and fusion noise enhancement adjustment value are output.

[0052] This example improves the dynamic fusion control capability of the fire extinguishing system by constructing a verification and optimization mechanism for the stability of mixed gas diffusion. It increases the uniform distribution speed of gas cloud in the high-temperature pyrolysis scenario of oil-immersed transformers, drives the Kalman filter to perform multi-dimensional noise suppression enhancement calculation, optimizes the signal-to-noise ratio of dynamic monitoring of oil mist-fire extinguishing gas mixing process, improves the convergence rate of gas stabilization time error, avoids the risk of reignition caused by the settling delay of extinguishing agent, and links the monitoring module to establish a real-time correction feedback link for the gas concentration distribution heat map.

[0053] like Figure 5 The diagram shows the structure of a multi-source data fusion system based on data analysis provided in an embodiment of the present invention. The multi-source data fusion system includes: a real-time fusion analysis module, a dynamic fusion adaptability analysis module, and an interactive fusion effectiveness analysis module. The real-time fusion analysis module is used by the fire monitoring and management platform to receive fire information data sent by IoT terminals and perform real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform based on the acquired real-time fusion data to determine whether sampling sensor frequency optimization is required. The dynamic fusion adaptability analysis module is used to perform dynamic adaptive analysis on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether sensing fusion parameter optimization is required. The interactive fusion effectiveness analysis module is used to perform interactive effectiveness analysis on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether interactive fusion noise optimization is required.

[0054] In this embodiment, the real-time analysis module dynamically adjusts the transmission efficiency of the alarm equipment, optimizes the sampling frequency, and shortens the latency to ensure that the fire signal and the platform are linked in real time; the dynamic fusion adaptive analysis module calibrates the parameters of the water spray equipment based on environmental variables to improve the accuracy of fire cooling; the interactive fusion effectiveness analysis module optimizes collaborative noise suppression through gas diffusion characteristic analysis; the modules cooperate with each other to jointly improve the efficiency of fire data fusion.

[0055] In summary, this invention, through real-time analysis of real-time fused data, sensing fused data, and the stabilization time of the mixed gas, triggers optimization of sampling sensing frequency, sensing fusion parameters, and interactive fusion noise, respectively. This allows fire sensors to detect abnormal information and transmit it to relevant fire equipment in real time. The real-time fused data is compared and corrected with the database to generate accurate scores. By coupling and quantifying real-time performance, the accuracy of delay assessment is improved. Sensing fusion parameters are optimized using sensing fusion data to dynamically compensate for delays caused by responses. The stabilization time of the mixed gas is used to evaluate the effectiveness of the interaction. This results in near-zero delay in receiving fire information in substations, near-zero equipment operation errors, and enhanced safety.

[0056] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0060] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0061] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A multi-source data fusion method based on data analysis, characterized in that, Includes the following steps: A1. The fire monitoring and management platform receives fire information data sent by IoT terminals. Based on the acquired real-time fusion data, it performs real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform to determine whether to optimize the sampling sensor frequency. The sampling sensor frequency optimization means improving the real-time synchronization of fire information data fusion by adjusting the checkpoint interval and node density. A2. If the synchronization fusion is deemed qualified, the platform performs dynamic adaptive analysis on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether to optimize the sensing fusion parameters. The sensing fusion parameter optimization means improving the effectiveness of fire information data sensing fusion by compensating for the window period caused by the sensing fusion response delay. A3. If the perception fusion is deemed qualified, the interaction effectiveness analysis of the fusion process between the fire extinguishing equipment and the fire monitoring and management platform is performed based on the obtained stable duration of the mixed gas to determine whether to perform interaction fusion noise optimization. The interaction fusion noise optimization means improving the interaction fusion effectiveness of fire information data by reducing noise interference in the interaction fusion process. The step of performing real-time performance analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform based on the acquired real-time fusion data includes: comparing the acquired real-time fusion data with the preset real-time fusion data in the database, and simultaneously performing correction processing based on the real-time fusion data correction value to obtain a real-time fusion data score; coupling the data coverage rate score, network jitter rate score, and electromagnetic interference intensity score to obtain a data fusion real-time performance impact index; the real-time fusion data includes data coverage rate, network jitter rate, and electromagnetic interference intensity, and the data fusion real-time performance impact index represents the quantitative data on the degree of influence of the real-time fusion data on the fusion real-time performance between the automatic fire alarm equipment and the fire monitoring and management platform; The specific steps for determining whether to optimize the sampling sensor frequency are as follows: compare the acquired data fusion real-time impact index with the preset data fusion real-time impact index in the database; if the acquired data fusion real-time impact index is greater than the preset data fusion real-time impact index in the database, it is recorded as data synchronization failure and sampling sensor frequency optimization is performed; if the acquired data fusion real-time impact index is not greater than the preset data fusion real-time impact index in the database, it is recorded as data synchronization success and dynamic adaptive analysis is performed. The method for dynamically adapting the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensor fusion data includes the following steps: Integrating the acquired sensor fusion data within a defined sensor fusion time interval to obtain the sensor fusion data area; and simultaneously coupling the sensor fusion data area to obtain a sensor fusion impact index. The sensor fusion time interval represents a unit time period used to analyze the sensor fusion efficiency. The integral processing represents the quantification of the cumulative changes in the sensor fusion data within the sensor fusion time interval. The sensor fusion data includes changes in temperature rise rate, combustion product smoke concentration, and smoke diffusion rate. The sensor fusion data area includes the areas of temperature rise rate changes, combustion product smoke concentration changes, and smoke diffusion rate changes. The sensor fusion impact index represents the quantified data on the degree of dynamic adaptability of the sensor fusion data in the fusion process between the transformer water spray equipment and the fire monitoring and management platform. The specific steps for determining whether to perform perception fusion parameter optimization are as follows: The obtained perception fusion impact index is compared with the preset perception fusion impact index in the database. If the obtained perception fusion impact index is not greater than the preset perception fusion impact index in the database, it is recorded as qualified perception fusion and interactive effectiveness analysis is performed. If the obtained perception fusion impact index is greater than the preset perception fusion impact index in the database, it is recorded as unqualified perception fusion and perception fusion parameter optimization is performed. The specific steps for performing interactive effectiveness analysis on the fusion process between the fire extinguishing equipment and the fire monitoring management platform based on the obtained mixed gas stabilization time to determine whether to perform interactive fusion noise optimization are as follows: The obtained mixed gas stabilization time is compared with the preset mixed gas stabilization time in the database. If the obtained mixed gas stabilization time is greater than the preset mixed gas stabilization time in the database, it is recorded as unstable adjustment and interactive fusion parameter optimization is performed. If the obtained mixed gas stabilization time is not greater than the preset mixed gas stabilization time in the database, it is recorded as stable adjustment and interactive effectiveness analysis is completed. The specific steps for optimizing the interactive fusion parameters are as follows: First, perform interactive fusion delay optimization: The obtained deviations in the stable duration of the mixed gas and the deviations in the release rate of the fire-fighting gas are proportionally processed with the preset deviations in the stable duration of the mixed gas and the release rate of the fire-fighting gas in the database, respectively. The results of the proportional processing are summed and averaged to obtain the interactive fusion delay correction value, which is used to correct the delay generated by the gas diffusion device, promote faster and more uniform distribution of the fire-fighting gas, and improve the response speed of the gas diffusion device. After the interactive fusion delay correction, it is determined whether the newly obtained stable duration of the mixed gas is not greater than the preset stable duration of the mixed gas in the database. If so, the first step is completed and the interactive effectiveness analysis is performed; otherwise, the second step is performed. The second step is to optimize the fusion noise: based on the deviation of the mixed gas stabilization time after the interactive fusion delay correction, the fusion noise enhancement value is mapped in the database to improve the noise amplitude suppression effect of the corresponding noise covariance matrix in the Kalman filter algorithm; after the interactive fusion parameter optimization, if the re-acquired mixed gas stabilization time is not greater than the preset mixed gas stabilization time in the database, the second step is completed and the interactive effectiveness analysis is completed; otherwise, the number of interactive fusion parameter optimizations is judged; the specific steps for judging the number of interactive fusion parameter optimizations are as follows: if the number of interactive fusion parameter optimizations is within the specified number of interactive fusion parameter optimizations, the first step is returned; otherwise, an interactive fusion warning is issued.

2. The multi-source data fusion method based on data analysis as described in claim 1, characterized in that, The specific steps for optimizing the sampling sensor frequency are as follows: First, optimize the checkpoint interval: The acquired deviations of the real-time impact index of data fusion and the alarm response time are proportionally processed against preset deviations of the real-time impact index of data fusion and the alarm response time in the database. The results of the proportional processing are summed and averaged to obtain a preliminary checkpoint interval value. This preliminary value is used to prompt the scheduler to adjust the checkpoint interval based on the acquired preliminary value to suppress glitches in the signal transmission band. After optimizing the sampling sensor frequency, it is determined whether the newly acquired deviation of the real-time impact index of data fusion is not greater than the preset deviation of the real-time impact index of data fusion in the database. If so, the first step is completed. Perform dynamic adaptive analysis; otherwise, proceed to step two. Step two involves node density optimization: based on the newly acquired node density reserve value, the topology self-organizing engine adjusts the node density to reduce the probability of phase distortion. After optimizing the sampling sensor frequency, if the newly acquired data fusion real-time impact index is not greater than the preset data fusion real-time impact index in the database, then step two is completed and dynamic adaptive analysis is performed; otherwise, a judgment is made regarding the number of sampling sensor frequency optimization attempts. The specific steps for judging the number of sampling sensor frequency optimization attempts are: if the number of sampling sensor frequency optimization attempts is within the specified number, then return to step one; otherwise, a sampling sensor frequency warning is issued.

3. The multi-source data fusion method based on data analysis as described in claim 1, characterized in that, The specific steps for optimizing the perception fusion parameters are as follows: First, perform data alignment correction: Obtain the perception fusion impact index deviation and inference delay time deviation, respectively, and perform proportional processing with the preset perception fusion impact index deviation and inference delay time deviation in the database. Sum and average the results after proportional processing to obtain the data alignment correction value, which is used to indicate the time loss caused by the injection controller due to the perception fusion response delay. The inference delay time represents the delay time generated by the smoke sensor for smoke recognition during combustion. After data alignment correction, determine whether the newly acquired perception fusion impact index is not greater than the preset perception fusion impact index in the database. If so, complete the first step. The first step involves performing an interaction effectiveness analysis; otherwise, it indicates an overloaded response delay, and the second step is initiated. The second step involves compensation for perception fusion delay: increasing the agent injection flow rate based on the newly acquired agent injection flow rate adjustment value to improve the control effectiveness of the fire in its initial stages. After optimizing the perception fusion parameters, if the newly acquired perception fusion impact index is not greater than the preset perception fusion impact index in the database, the second step is completed, and an interaction effectiveness analysis is performed; otherwise, a judgment is made regarding the number of times the perception fusion parameters have been optimized. The specific steps for judging the number of times the perception fusion parameters have been optimized are as follows: if the number of times the perception fusion parameters have been optimized is within the specified number of times, the first step is returned; otherwise, a perception fusion warning is issued.

4. A system applying the multi-source data fusion method based on data analysis as described in any one of claims 1-3, characterized in that, The system includes a real-time fusion analysis module, a dynamic fusion adaptability analysis module, and an interactive fusion effectiveness analysis module. The real-time fusion analysis module is used by the fire monitoring and management platform to receive fire information data sent by IoT terminals and perform real-time analysis on the fusion process between the automatic fire alarm equipment and the fire monitoring and management platform based on the acquired real-time fusion data to determine whether sampling sensor frequency optimization is necessary. The dynamic fusion adaptability analysis module is used to perform dynamic adaptive analysis on the fusion process between the transformer water spray equipment and the fire monitoring and management platform based on the acquired sensing fusion data to determine whether sensing fusion parameter optimization is necessary. The interactive fusion effectiveness analysis module is used to perform interactive effectiveness analysis on the fusion process between the fire extinguishing equipment and the fire monitoring and management platform based on the acquired mixed gas stabilization time to determine whether interactive fusion noise optimization is necessary.

Images