CN-121724445-B - Chemical laboratory risk early warning method and system based on multi-mode data fusion algorithm

CN121724445BCN 121724445 BCN121724445 BCN 121724445BCN-121724445-B

Abstract

The invention relates to the technical field of intelligent safety monitoring, in particular to a chemical laboratory risk early warning method and a chemical laboratory risk early warning system based on a multi-mode data fusion algorithm, which realize deep insight on dynamic change of laboratory environment by calculating spatial gradients of multi-mode data such as gas concentration, temperature and the like and analyzing coupling relation and time delay between the gradients, do not depend on isolated single-point threshold judgment any more, evaluate stability according to cooperative phases of thermal diffusion and wet flow processes, according to the method, potential unstable states caused by the combined action of multiple factors can be effectively identified, even if all single parameters do not reach the alarm limit value, risk sprouting can be captured in advance, meanwhile, through aggregation analysis on the unstable areas adjacent in space, the risk spreading trend and the influence range can be revealed, so that more accurate early warning with forward looking is generated, and false alarm caused by non-dangerous instantaneous fluctuation is remarkably reduced.

Inventors

LIU QIAN
YANG ZHENDONG
LIANG BO
SUN HAO
Yuan Fanyu
NING XIAOPENG

Assignees

四川安信科创科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260213

Claims (6)

1. The chemical laboratory risk early warning method based on the multi-mode data fusion algorithm is characterized by comprising the following steps of: s1, acquiring gas concentration, temperature, humidity and air flow speed data at a designated position in a chemical laboratory, respectively calculating gradients of each type of data among adjacent monitoring points in respective preset three-dimensional space, and establishing a multi-mode synchronous monitoring data set; S2, calculating a thermal diffusion coupling parameter and a wet flow coupling parameter according to the gradient direction of each type of data in the multi-mode synchronous monitoring data set, determining a target space position with obvious coupling, and extracting local coupling change data at the target space position; S3, analyzing the thermal diffusion time delay of the gas concentration and the temperature change at the target space position in the local coupling change data, and carrying out fusion matching on the wet flow time delay of the humidity change and the air flow speed change to obtain time sequence phase coupling data; s4, determining a cooperative stable phase region or a potential unstable phase region of thermal diffusion and wet flow according to the time sequence phase coupling data, and integrating the cooperative stable phase region or the potential unstable phase region into a multi-mode risk distribution data set; s5, performing regional aggregation analysis processing on the multi-mode risk distribution data set along laboratory space coordinates to generate a risk early warning result; the step S2 specifically comprises the following steps: S201, extracting direction parameters of concentration gradient, temperature gradient, humidity gradient and air flow velocity gradient in the multi-mode synchronous monitoring data set, calculating an included angle between the temperature gradient and the concentration gradient direction aiming at the same space coordinate position to obtain a thermal diffusion coupling parameter, calculating an included angle between the humidity gradient and the air flow velocity gradient direction to obtain a wet flow coupling parameter, and establishing a thermal wet coupling parameter set; S202, calling a preset thermal diffusion coupling threshold value and a wet flow coupling threshold value, and based on the comparison of the thermal and wet coupling parameter sets, screening the space positions with the thermal diffusion coupling parameter smaller than the thermal diffusion coupling threshold value and the wet flow coupling parameter smaller than the wet flow coupling threshold value as the obvious target space positions; S203, extracting concentration change rate, temperature change rate, humidity change rate and airflow speed change rate corresponding to the target space position with obvious coupling from the multi-mode synchronous monitoring data set, and establishing local coupling change data; The step S3 specifically comprises the following steps: S301, calling a target space position with obvious coupling in the local coupling change data, and calculating time delay between a gas concentration change rate sequence and a temperature change rate sequence of the target space position to obtain thermal diffusion time delay; S302, calling a target space position with obvious coupling in the local coupling change data, calculating time delay between a humidity change rate sequence and an airflow speed change rate sequence of the target space position, and obtaining wet flow time delay; S303, invoking a time sequence corresponding to the thermal diffusion time delay and the wet flow time delay to be fused and matched with a space coordinate, and establishing time sequence phase coupling data; The step S4 specifically comprises the following steps: s401, comparing the thermal diffusion time delay and the wet flow time delay in the time sequence phase coupling data with respective thermal diffusion threshold values and preset wet flow threshold values respectively to generate a phase region calibration result; S402, searching and extracting concentration gradient data, temperature gradient data, humidity gradient data and air flow speed gradient data corresponding to the positions from the multi-mode synchronous monitoring data set according to the calibrated spatial positions in the phase region calibration result, and establishing a target region gradient data set; S403, associating with the common spatial position information in the two data of the target region gradient data set according to the phase region calibration result, and establishing a multi-mode risk distribution data set; The step S5 specifically comprises the following steps: S501, extracting potential unstable phase areas in the same time window from the multi-mode risk distribution data set, calculating a spatial adjacent relation between each potential unstable phase area, judging the aggregation association degree of adjacent areas according to the coordinate distance and the gradient amplitude weight, and establishing a spatial aggregation association area set; s502, acquiring heat diffusion time delay and gradient data in the space aggregation related region set according to the multi-mode risk distribution data set, calculating average heat diffusion phase difference and average wet flow phase difference of each aggregation region in the space aggregation related region set, and judging consistency of phase difference change trend and gradient change direction in the aggregation region in the space aggregation related region set to generate an aggregation region risk assessment result; And S503, comparing the average thermal diffusion phase difference and the average wet flow phase difference of the aggregation region risk assessment result with respective risk judgment thresholds, and generating records on corresponding aggregation region space coordinates when any average phase difference exceeds the corresponding risk judgment threshold, so as to establish a risk early warning result.
2. The chemical laboratory risk early warning method based on a multi-modal data fusion algorithm according to claim 1, wherein the multi-modal synchronous monitoring dataset comprises concentration gradient distribution, temperature gradient distribution, humidity gradient distribution and airflow velocity gradient distribution, the local coupling change data specifically comprise concentration change rate, temperature change rate, humidity change rate and airflow velocity change rate at a target space position, the time sequence phase coupling data comprise thermal diffusion time delay, wet flow time delay, space coordinates and time markers, the multi-modal risk distribution dataset comprises phase difference parameters corresponding to synergistically stable phase region information, potentially unstable phase region information and regions, and the risk early warning result specifically comprises a risk space position and a risk level.
3. The chemical laboratory risk early warning method based on the multi-modal data fusion algorithm according to claim 1, wherein the step of S1 specifically comprises: S101, acquiring original monitoring data acquired by a gas concentration sensor, a temperature sensor, a humidity sensor and an air flow speed sensor at a designated position in a chemical laboratory, executing synchronous alignment according to a sensor time stamp, recording the gas concentration, the temperature, the humidity and the air flow speed at the same time point, correlating with three-dimensional space coordinates calibrated by a sensor layout, and generating a monitoring data space coordinate mapping set; s102, determining adjacent monitoring points in a three-dimensional space coordinate system based on the monitoring data space coordinate mapping set, calling the coordinate data of the adjacent monitoring points to calculate Euclidean space distance between the two points, acquiring gas concentration difference, temperature difference, humidity difference and air flow speed difference of the adjacent monitoring points under the same time stamp, and dividing each difference by the corresponding space distance to obtain a multi-dimensional monitoring data space gradient value; and S103, calling the monitoring data space coordinate mapping set and the multi-dimensional monitoring data space gradient value, and establishing a joint index for the gas concentration, the temperature, the humidity, the gas flow speed and the corresponding gradient data under a unified coordinate system to obtain a multi-mode synchronous monitoring data set.
4. The method for chemical laboratory risk early warning based on a multimodal data fusion algorithm according to claim 1, characterized in that the process of comparing the respective thermal diffusion threshold with a preset wet flow threshold comprises: Traversing each data record in the sequential phase-coupled data and extracting a thermal diffusion time delay and a wet flow time delay from the data record; Performing numerical value judgment on the extracted thermal diffusion time delay and a thermal diffusion threshold value, and simultaneously performing numerical value judgment on the extracted wet flow time delay and a preset wet flow threshold value; if the thermal diffusion time delay is smaller than the thermal diffusion threshold and the wet flow time delay is smaller than the preset wet flow threshold, calibrating the spatial position related to the data record as a collaborative stable phase region; If the thermal diffusion time delay is greater than or equal to the thermal diffusion threshold, or the wet flow time delay is greater than or equal to the preset wet flow threshold, the spatial position associated with the data record is marked as a potential unstable phase region.
5. The chemical laboratory risk early warning method based on the multi-modal data fusion algorithm according to claim 1, wherein the process of determining the consistency of the phase difference change trend and the gradient change direction in the centralized aggregation area of the spatial aggregation association area comprises: Traversing each aggregation region in the spatial aggregation association region set, and extracting thermal diffusion time delay, wet flow time delay, concentration gradient, temperature gradient, humidity gradient and airflow velocity gradient data of continuous time sequences in the aggregation region; Calculating a variation trend vector of the thermal diffusion time delay and the wet flow time delay according to the continuous time sequence; according to the continuous time sequence, respectively calculating the comprehensive change direction vector of the temperature gradient and the concentration gradient and the comprehensive change direction vector of the humidity gradient and the air flow velocity gradient; calculating an included angle between a change trend vector of the thermal diffusion time delay and a comprehensive change direction vector of the temperature gradient and the concentration gradient, and comparing the included angle with a preset consistency angle threshold; Meanwhile, calculating an included angle between a change trend vector of the wet flow time delay and a comprehensive change direction vector of the humidity gradient and the airflow velocity gradient, and comparing the included angle with a consistency angle threshold; and when the two calculated included angles are smaller than the consistency angle threshold value, judging that the phase difference change trend is consistent with the gradient change direction.
6. A chemical laboratory risk early warning system based on a multi-modal data fusion algorithm, characterized in that the system is used for implementing the chemical laboratory risk early warning method based on the multi-modal data fusion algorithm according to any one of claims 1 to 5, and the system comprises: The data gradiometer module is used for acquiring gas concentration, temperature, humidity and air flow speed data at a designated position in the chemical laboratory, respectively calculating the gradient of each type of data between adjacent monitoring points in a respective preset three-dimensional space, and establishing a multi-mode synchronous monitoring data set; the space coupling positioning module is used for calculating a thermal diffusion coupling parameter and a wet flow coupling parameter according to the gradient direction of each type of data in the multi-mode synchronous monitoring data set, determining a target space position with obvious coupling, and extracting local coupling change data at the target space position; The phase coupling analysis module is used for analyzing the thermal diffusion time delay of the gas concentration and the temperature change at the target space position in the local coupling change data and the wet flow time delay of the humidity change and the air flow speed change, and performing fusion matching to obtain time sequence phase coupling data; The risk area identification module is used for determining a cooperative stable phase area or a potential unstable phase area of the thermal diffusion and the wet flow according to the time sequence phase coupling data and integrating the cooperative stable phase area or the potential unstable phase area into a multi-mode risk distribution data set; And the risk aggregation early warning module is used for executing regional aggregation analysis processing along the laboratory space coordinates aiming at the multi-mode risk distribution data set to generate a risk early warning result.

Description

Chemical laboratory risk early warning method and system based on multi-mode data fusion algorithm Technical Field The invention relates to the technical field of intelligent safety monitoring, in particular to a chemical laboratory risk early warning method and system based on a multi-mode data fusion algorithm. Background The technical field of intelligent safety monitoring relates to the use of advanced technical means to monitor and pre-warn potential risks in various environments in real time, so as to ensure the safety of personnel and equipment. In this field, a plurality of technical directions are covered, including but not limited to applications such as video monitoring, sensor networks, data analysis, artificial intelligence, machine learning, etc., to realize real-time detection and early warning of potential safety hazards. The chemical laboratory risk early warning method is to predict and early warn potential risks in a chemical laboratory by using a traditional monitoring technology. Such methods typically collect environmental parameters in a laboratory in real time by setting up fixed monitoring equipment such as temperature, humidity, gas concentration sensors, etc. Based on the collected data, the system will issue an alarm to alert laboratory staff to take action in time. The existing chemical laboratory risk early warning method mainly relies on independent monitoring of single environmental parameters, the operation mode is limited by directly comparing acquired discrete data points with preset static thresholds, the mechanism ignores inherent dynamic correlations among different physical quantities, and cannot deeply analyze interactions such as heat changes and gas diffusion in space, so that the phenomenon of poor performance is caused when the instantaneous fluctuation caused by normal operation is distinguished from the continuous evolution of real potential risks, the loss of judging capability is not only easy to trigger false alarms due to local harmless disturbance, the reliability of early warning is reduced, but also more critical, for some slowly developed but dangerous abnormal working conditions, risks cannot be recognized in advance due to the fact that no single index touches an alarm line, and finally the prospective and accuracy of early warning are weakened. Disclosure of Invention In order to achieve the purpose, the chemical laboratory risk early warning method based on the multi-mode data fusion algorithm comprises the following steps: s1, acquiring gas concentration, temperature, humidity and air flow speed data at a designated position in a chemical laboratory, respectively calculating gradients of each type of data among adjacent monitoring points in respective preset three-dimensional space, and establishing a multi-mode synchronous monitoring data set; S2, calculating a thermal diffusion coupling parameter and a wet flow coupling parameter according to the gradient direction of each type of data in the multi-mode synchronous monitoring data set, determining a target space position with obvious coupling, and extracting local coupling change data at the target space position; S3, analyzing the thermal diffusion time delay of the gas concentration and the temperature change at the target space position in the local coupling change data, and carrying out fusion matching on the wet flow time delay of the humidity change and the air flow speed change to obtain time sequence phase coupling data; s4, determining a cooperative stable phase region or a potential unstable phase region of thermal diffusion and wet flow according to the time sequence phase coupling data, and integrating the cooperative stable phase region or the potential unstable phase region into a multi-mode risk distribution data set; and S5, carrying out regional aggregation analysis processing on the multi-mode risk distribution data set along the laboratory space coordinates to generate a risk early warning result. As a further scheme of the invention, the multi-mode synchronous monitoring data set comprises concentration gradient distribution, temperature gradient distribution, humidity gradient distribution and airflow speed gradient distribution, the local coupling change data specifically comprise concentration change rate, temperature change rate, humidity change rate and airflow speed change rate at a target space position, the time sequence phase coupling data comprise thermal diffusion time delay, wet flow time delay, space coordinates and time marks, the multi-mode risk distribution data set comprises phase difference parameters corresponding to a synergistic stable phase region information, potential unstable phase region information and region, and the risk early warning result specifically refers to a risk space position and a risk grade. As a further scheme of the invention, the step S1 specifically comprises the following steps: S101, acquiring original monitoring data acquired by a