Limited space operation risk identification and early warning method, device, equipment and medium

Abstract

The application provides a limited space operation risk identification and early warning method, device, equipment and medium, and relates to the technical field of safety monitoring. The method comprises the following steps: calculating a difference vector of a real-time environment state vector of a target area and a reference environment state vector; wherein the environment state vector comprises a toxic and harmful gas concentration, a water level and a water flow rate; calculating a Mahalanobis distance based on the difference vector and a reference covariance matrix as a joint abnormality index at the current moment; updating an integral accumulation value based on the toxic and harmful gas concentration of the target area at the current moment, and calculating a comprehensive risk index at the current moment based on the comprehensive risk index at the last moment, the updated integral accumulation value and the joint abnormality index at the current moment; and performing risk early warning on the target area based on the comprehensive risk index at the current moment. The application can effectively improve the accuracy and timeliness of the operation risk identification and early warning of the limited space.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology, and in particular to a method, device, equipment and medium for identifying and warning of operational risks in confined spaces. Background Technology

[0002] Construction work in confined spaces such as underground utility tunnels and sewage wells presents high-risk conditions due to poor ventilation and susceptibility to complex hydrological conditions. To ensure personnel safety, construction companies typically deploy real-time monitoring systems within these confined spaces, relying on sensors to continuously collect and dynamically monitor key environmental parameters such as the concentration of toxic and harmful gases, water levels, and water flow rates.

[0003] However, currently widely used monitoring technologies typically set alarm thresholds independently for each environmental parameter. The system only triggers an alarm when the actual measured value of a parameter exceeds its preset threshold. This monitoring method has significant limitations: First, it only issues an alarm after an environmental parameter has reached or exceeded a danger threshold, by which time the danger has often already occurred, leaving extremely limited time for on-site personnel to safely evacuate. Second, this independent judgment method ignores the inherent correlations between different environmental parameters. For example, sudden changes in water level often disturb sediments and are accompanied by the release of harmful gases, and changes in water flow velocity may indicate drastic changes in hydrological conditions. These early signs of multi-factor synergistic changes are difficult to effectively identify under independent threshold monitoring, making it difficult to capture abnormal evolution trends in environmental conditions before a danger occurs, thus failing to provide sufficient response time for on-site personnel. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for identifying and warning of operational risks in confined spaces, in order to solve the problem of low accuracy and timeliness in identifying and warning of operational risks in confined spaces.

[0005] In a first aspect, embodiments of the present invention provide a method for identifying and warning of operational risks in confined spaces, comprising: Calculate the difference vector between the real-time environmental state vector of the target area and the baseline environmental state vector; whereby the environmental state vector includes the concentration of toxic and harmful gases, water level, and water flow velocity; The Mahalanobis distance is calculated based on the difference vector and the benchmark covariance matrix, and used as the joint anomaly index at the current moment. The integral cumulative value is updated based on the concentration of toxic and harmful gases in the target area at the current time, and the comprehensive risk index at the current time is calculated based on the comprehensive risk index at the previous time, the updated integral cumulative value, and the joint anomaly index at the current time. Risk warnings are issued for the target area based on the comprehensive risk index at the current moment.

[0006] In one possible implementation, before calculating the difference between the real-time environmental state vector of the target region and the reference environmental state vector, the following steps are also included: Calculate the mathematical expectations of the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, and form a benchmark mean vector; Based on the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, calculate the covariance values ​​between parameter sequences of each dimension. A positive definite symmetric square matrix representing the joint distribution characteristics is constructed based on the values ​​of each covariance, serving as the benchmark covariance matrix.

[0007] In one possible implementation, the Mahalanobis distance is calculated based on the difference vector and the baseline covariance matrix, serving as the joint anomaly index at the current time, including: Perform matrix multiplication operations sequentially on the transpose of the difference vector, the reference covariance matrix, and the difference vector to obtain a quadratic scalar that characterizes the degree of multi-parameter coupling distortion. The joint anomaly index at the current moment is obtained by taking the square root of the quadratic scalar.

[0008] In one possible implementation, the integral cumulative value is updated based on the concentration of toxic and harmful gases in the target area at the current moment, including: Update the gas concentration sequence based on the current concentration of toxic and harmful gases in the target area; The updated gas concentration sequence is convolved with a preset monotonically decreasing time decay weight function to obtain the updated cumulative integral value.

[0009] In one possible implementation, the comprehensive risk index at the current moment is calculated based on the comprehensive risk index of the previous moment, the updated cumulative integral value, and the joint anomaly index at the current moment, including: The instantaneous evolution rate of the environmental state is obtained by performing first-order difference calculation on the joint anomaly index of the current moment and multiple historical moments. The instantaneous evolution rate is input into a preset mapping function to obtain dynamically updated weight coefficients; Using weighting coefficients, the comprehensive risk warning index of the previous time step, the updated cumulative integral value, and the joint anomaly index of the current time step are weighted and summed to obtain the comprehensive risk index of the current time step.

[0010] In one possible implementation, risk warnings are issued for the target area based on the comprehensive risk index at the current moment, including: The risk level of the target area at the current moment is obtained by comparing the comprehensive risk index at the current moment with the preset grading threshold range. Based on the risk level of the target area at the current moment, warnings are issued through corresponding warning methods; these methods include controlling ground displays and / or mobile terminals to output warning prompts, triggering on-site audible and visual alarm devices, and generating evacuation control signals.

[0011] Secondly, embodiments of the present invention provide a confined space operation risk identification and early warning device, comprising: The difference calculation module is used to calculate the difference vector between the real-time environmental state vector of the target area and the baseline environmental state vector; wherein, the environmental state vector includes the concentration of toxic and harmful gases, water level and water flow velocity; The anomaly index calculation module is used to calculate the Mahalanobis distance based on the difference vector and the benchmark covariance matrix, which serves as the joint anomaly index at the current moment. The risk index calculation module is used to update the cumulative integral value based on the concentration of toxic and harmful gases in the target area at the current time, and to calculate the comprehensive risk index at the current time based on the comprehensive risk index at the previous time, the updated cumulative integral value, and the joint anomaly index at the current time. The risk warning module is used to issue risk warnings to the target area based on the comprehensive risk index at the current moment.

[0012] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect or any possible implementation thereof.

[0013] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect or any possible implementation thereof.

[0014] Fifthly, embodiments of the present invention provide a computer program product, including a computer program that, when executed by a processor, implements the method described in the first aspect or any possible implementation thereof.

[0015] The confined space operation risk identification and early warning method, device, equipment, and medium provided in this invention construct a three-dimensional environmental state vector containing toxic and harmful gas concentrations, water levels, and water flow velocities. Based on the degree of vector deviation from the baseline state and combined with the covariance relationship between multiple parameters, Mahalanobis distance is calculated, enabling a comprehensive quantification of the joint distortion degree of multiple environmental parameters. Compared to the discrimination method based on independent thresholds for single parameters, this method can identify potential hazards in advance when a single parameter has not yet exceeded the standard but the coordinated changes of multiple parameters have already shown an abnormal trend. Simultaneously, by introducing a gas concentration integral cumulative value based on time decay weights, it can reflect the hidden harm suffered by workers due to prolonged exposure to low-concentration harmful gas environments. Combined with the risk state at the previous moment for dynamic weighted fusion, the risk assessment considers both sudden changes and cumulative effects, effectively improving the accuracy and timeliness of confined space operation risk identification and early warning. Finally, by triggering corresponding response actions under different risk levels through a graded early warning strategy, it helps to provide more adequate safety guarantees for confined space operations while ensuring normal construction order. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the implementation of the confined space operation risk identification and early warning method provided in this embodiment of the invention. Figure 2 This is a schematic diagram of the structure of the confined space operation risk identification and early warning device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0017] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] See Figure 1 The flowchart illustrating the implementation of the confined space operation risk identification and early warning method provided in this embodiment of the invention is described in detail below: Step 101: Calculate the difference vector between the real-time environmental state vector of the target area and the baseline environmental state vector; wherein, the environmental state vector includes the concentration of toxic and harmful gases, water level and water flow velocity.

[0019] In this embodiment, the environmental state vector is a three-dimensional vector composed of three parameters: toxic and harmful gas concentration, water level, and water flow velocity. These three parameters are the most critical environmental indicators in the safety assessment of confined space operations. The real-time environmental state vector reflects the actual environmental condition of the target area at the current moment, and its data comes from monitoring equipment such as gas sensors, ultrasonic level gauges, and flow meters deployed within the confined space. The baseline environmental state vector represents the standard environmental characteristics of the target area under normal and safe conditions, and its value is calculated by collecting historical data during the system initialization phase. When a deviation occurs between the real-time environmental state vector and the baseline environmental state vector, it indicates that the current environment has changed relative to the normal state. Subtracting the components of corresponding dimensions of the two vectors yields the difference vector, which quantifies the direction and degree of deviation of the current environment from the baseline value in each parameter dimension. The significance of this calculation lies in transforming the differences in multi-dimensional environmental parameters into a unified vector form, laying the data foundation for subsequent multi-parameter joint analysis, enabling the system to grasp the changes in environmental state as a whole, rather than observing a single parameter in isolation.

[0020] Step 102: Calculate the Mahalanobis distance based on the difference vector and the benchmark covariance matrix, which serves as the joint anomaly index at the current moment.

[0021] In this embodiment, Mahalanobis distance is a mathematical metric used to measure the distance between a sample vector and a known distribution. Its key difference from ordinary Euclidean distance lies in the introduction of a covariance matrix to characterize the correlation between dimensions. In confined space operating environments, the concentration of toxic and harmful gases, water level, and water flow velocity are not independent but have an inherent physical coupling relationship. For example, a rise in water level may wash away sediment and agitate the bottom water, causing dissolved harmful gases to be released into the air; simultaneously, an increase in water flow velocity may exacerbate this release process. The baseline covariance matrix is ​​calculated based on historical data during the system initialization phase. It accurately records the joint distribution characteristics of these three parameters under normal operating conditions and their covariant relationships. Substituting the difference vector into the Mahalanobis distance calculation formula essentially determines the degree to which the multi-parameter combination deviates from the normal joint distribution at the current moment. When the three parameters change synchronously according to the normal covariant relationship, even if the absolute values ​​of each parameter fluctuate, the Mahalanobis distance will not increase significantly; only when the linkage between the parameters is broken or the overall deviation reaches an abnormal level will the Mahalanobis distance increase significantly. Therefore, this distance value comprehensively reflects the degree of multi-parameter coupling distortion and is suitable as a joint anomaly index to characterize the current environmental anomaly, providing a quantitative basis for subsequent risk assessment.

[0022] Step 103: Update the cumulative integral value based on the concentration of toxic and harmful gases in the target area at the current time, and calculate the comprehensive risk index at the current time based on the comprehensive risk index at the previous time, the updated cumulative integral value, and the joint anomaly index at the current time.

[0023] In this embodiment, the cumulative integral value means the summation of toxic and harmful gas concentrations monitored over a continuous period of time, with a time decay effect. In confined space operations, personnel safety depends not only on the instantaneous gas concentration at the current moment but also on the cumulative dose from continuous exposure over a past period. Even if the current concentration does not reach an immediately harmful level, prolonged exposure to low-concentration harmful environments can cause latent damage to human health, which gradually worsens with prolonged exposure. Therefore, whenever gas concentration data for the current moment is obtained, the system incorporates it into the historical concentration sequence and updates the cumulative value through integral calculation, ensuring that the value always reflects the cumulative exposure status at the current moment. The comprehensive risk index is a comprehensive evaluation indicator that integrates multiple risk factors. Its calculation considers information from three dimensions: the risk level at the previous moment inherits the historical risk situation, avoiding drastic changes in risk assessment; the updated cumulative integral value reflects the chronic damage from long-term exposure; and the current joint anomaly index reflects the sudden risk of abrupt environmental changes. The combination of these three factors ensures that the comprehensive risk index will not frequently report false alarms due to instantaneous fluctuations, and can respond promptly when the environment continues to deteriorate or changes suddenly, thus providing a comprehensive and stable risk assessment for confined space operations.

[0024] Step 104: Issue a risk warning for the target area based on the comprehensive risk index at the current moment.

[0025] In this embodiment, risk warning is a crucial step in transforming the comprehensive risk index calculated in the aforementioned steps into specific safety assurance actions. The comprehensive risk index, as a continuous numerical value, represents the level of risk in the current confined space working environment. To convert it into a signal that can be used for decision-making or execution, the system pre-stores graded risk threshold ranges, dividing the numerical range of the comprehensive risk index into different risk level intervals. By determining which interval the current comprehensive risk index falls into, the system can determine the risk level of the target area at the current moment. Different risk levels correspond to different response strategies: at lower risk levels, the system pushes warning information to the ground display screen and the mobile terminals of on-site monitoring personnel, reminding them to pay attention to environmental changes. At this time, physical alarm devices are not triggered to avoid interfering with normal construction. At higher risk levels, the system automatically triggers the activation of on-site audible and visual alarm devices and sends evacuation control signals to the escape guidance equipment, achieving mandatory physical intervention. This graded warning method ensures that the system can take appropriate response measures under different risk levels, achieving both timely identification and notification of potential hazards and buying valuable evacuation time for on-site personnel when extreme dangers occur.

[0026] This invention constructs a three-dimensional environmental state vector containing toxic and harmful gas concentrations, water levels, and water flow velocities. Based on the degree of vector deviation from the baseline state and combined with the covariance relationship between multiple parameters, Mahalanobis distance is calculated, enabling a comprehensive quantification of the joint distortion degree of multiple environmental parameters. Compared to methods that rely on independent thresholds for single parameters, this approach can identify potential hazards in advance when a single parameter has not yet exceeded its limit, but the coordinated changes of multiple parameters have already shown an abnormal trend. Furthermore, by introducing a gas concentration integral accumulation value based on time decay weights, it can reflect the hidden harm suffered by workers due to prolonged exposure to low-concentration harmful gas environments. Combined with the risk status of the previous moment for dynamic weighted fusion, the risk assessment considers both sudden changes and cumulative effects, effectively improving the accuracy and timeliness of risk identification and early warning in confined spaces. Finally, a tiered early warning strategy triggers corresponding response actions at different risk levels, helping to provide more adequate safety guarantees for confined space operations while ensuring normal construction order.

[0027] In one possible implementation, before calculating the difference between the real-time environmental state vector of the target region and the reference environmental state vector, the following steps are also included: Calculate the mathematical expectations of the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, and form a benchmark mean vector; Based on the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, calculate the covariance values ​​between parameter sequences of each dimension. A positive definite symmetric square matrix representing the joint distribution characteristics is constructed based on the values ​​of each covariance, serving as the benchmark covariance matrix.

[0028] In this embodiment, the steps of constructing the baseline environmental state vector and the baseline covariance matrix are completed during the system initialization phase before the construction operation begins. The sensor array deployed within the limited space continuously collects environmental parameter sequences within a preset initialization time window; the data collected in parallel by each channel constitutes the historical data matrix. Due to the thermal noise of the sensor hardware itself and electromagnetic interference in long-distance communication links, the original collected data inevitably contains high-frequency noise components. Although these noises are small in amplitude, they can interfere with the accurate extraction of baseline features. Therefore, a filtering algorithm is needed to perform smoothing processing before calculating statistical features. Taking sliding window filtering as an example, for the first... Original acquisition sequence of each feature dimension With the preset window width Slide the slider on the timeline and take the mean of the data within the window as the smoothed output value for that moment. : The mathematical essence of this formula is a reflection of the past. The sampling points are averaged equally to form a low-pass filter, whose frequency response is in The presence of a zero point effectively suppresses random noise components above the cutoff frequency while retaining the gradually changing trend components reflecting the true physical state of the environment, thus reducing the negative impact of sensor jitter and communication interference on baseline feature extraction. After smoothing, the data sequence needs to be scanned by anomaly detection logic to remove outlier samples caused by transient disturbances such as sudden changes in downhole airflow or floating foreign objects on the water surface. Although these outliers are few in number, retaining them in the baseline data would significantly deviate the estimated values ​​of statistical features, reducing the sensitivity of subsequent anomaly detection. Specifically, a discrimination criterion based on local expected value and standard deviation can be used, where the expected value of the smoothed sequence within the local window is... The standard deviation is When a sampling point meets the following conditions: The sample point is determined to be an outlier caused by a transient perturbation and is forcibly removed from the sequence. The theoretical basis for this criterion is the 3... The criterion is that, under the assumption of a normal distribution, the probability of a sample falling outside three standard deviations is only 2.7 per thousand, which is a low-probability event. Therefore, identifying an outlier under this condition is statistically reliable. The final output is the cleaned, effective baseline data. Based on this, the parameter sequences are reorganized into a multidimensional matrix according to a fixed-dimensional order of toxic gas concentration, water level, and water flow velocity. Let the total number of effective baseline samples be... , No. The concentration of toxic and harmful gases in each sample was Water level is The water flow velocity is Then the formula for calculating the expected value of the corresponding dimension is: The physical meaning of these three formulas is to take the arithmetic mean of all samples in each dimension. In probability theory, this statistic is an unbiased estimate of the population expectation, and its value increases with the sample size. As the value increases, its estimation bias decreases. The rate converges to zero, therefore, when the initialization time window is long enough, the benchmark mean vector can accurately reflect the true feature center of the finite space under normal conditions. The system combines these three expected values ​​in a predetermined order to form the benchmark mean vector of the three-dimensional environmental state vector characterizing the safe and normal characteristics of the finite space. Simultaneously, the system iterates through and calculates the covariance values ​​between the parameter sequences of each feature dimension to quantify the internal physical correlation between toxic gas release, water level changes, and water flow fluctuations. The covariance matrix contains the [missing information - likely a specific element or component]. Line 1 The formula for calculating the elements of a column is: in and Representing the first In the nth sample The feature dimension and the first The specific values ​​of each feature dimension and This represents the expected value of the corresponding feature dimension. The denominator of this formula uses... Instead This is to achieve unbiased estimation, since the mean is unknown before calculating the covariance and needs to be estimated from samples, resulting in a loss of 1 degree of freedom. The sign of the off-diagonal elements reflects the direction of the cooperative change between the two parameters, and the magnitude reflects the strength of the cooperative change. These coupling relationships will be used as scaling factors in the metric space in the subsequent Mahalanobis distance calculation, enabling the system to distinguish between normal cooperative fluctuations and true anomalous deviations. After calculating the covariance elements across all dimensions, the system constructs a third-order positive definite symmetric matrix, which is used as the covariance matrix. This step extracts the characteristic distribution patterns of a finite space under a stable and risk-free state, establishing a quantitative comparison standard for the system to subsequently determine whether environmental data has undergone multi-parameter abnormal coupling distortion.

[0029] In one possible implementation, the Mahalanobis distance is calculated based on the difference vector and the baseline covariance matrix, serving as the joint anomaly index at the current time, including: Perform matrix multiplication operations sequentially on the transpose of the difference vector, the reference covariance matrix, and the difference vector to obtain a quadratic scalar that characterizes the degree of multi-parameter coupling distortion. The joint anomaly index at the current moment is obtained by taking the square root of the quadratic scalar.

[0030] In this embodiment, let the difference vector be... The baseline covariance matrix is Then the mathematical form of matrix chain multiplication is: This quadratic scalar Geometrically, it refers to a system with a reference mean vector as the origin and a covariance matrix as the origin. Define the squared length of the current deviation vector measured in the elliptic space of the metric tensor. Euclidean distance. The difference lies in the insertion of the formula. This achieves anisotropic scaling: in the eigenvector directions of the covariance matrix, the scaling factor is proportional to the reciprocal of the corresponding eigenvalue. This means that in parameter directions that are inherently more volatile under normal operating conditions, larger deviations are allowed without being considered abnormal; while in directions with smaller fluctuations, even small deviations will be scaled up. The magnitude of the difference is significantly amplified. This characteristic is highly consistent with the actual physical laws of confined space environments—for example, water levels fluctuate considerably during normal operations due to factors such as construction water discharge, while gas concentration changes are usually relatively small. If Euclidean distance is used for direct measurement, the large normal fluctuations in water level will dominate the distance value, thus obscuring the slight anomalies in gas concentration. In contrast, Mahalanobis distance... It can automatically compensate for different variance scales across dimensions, ensuring that the contribution of each parameter to the joint anomaly index matches its own physical fluctuation characteristics. However, under extreme conditions, water level and flow velocity may exhibit extremely high linear collinearity, leading to an imbalance in the covariance matrix. When the ratio of the largest eigenvalue to the smallest eigenvalue exceeds a preset condition number threshold, the matrix tends to be ill-conditioned, and direct inversion will cause numerical overflow. In this case, the system adds a perturbation regularization term to the main diagonal. ,in The preset minimal perturbation constant, Since the matrix is ​​an identity matrix, this operation reduces the condition number by translating the eigenvalue spectrum, thereby obtaining the generalized inverse matrix. And ensure the numerical stability of the quadratic form calculation. Obtain Then, perform the square root operation on it: This square root operation restores the dimensions from quadratic to a linear scale consistent with the original data space, thus enabling the joint anomaly index to... The numerical range is comparable to the original magnitude of each parameter. The Joint Anomaly Index is a comprehensive reflection of the degree of joint anomaly dispersion of toxic and harmful gases, water level, and water flow velocity at the current moment. The larger the value, the further the current environmental state deviates from the normal joint distribution.

[0031] In one possible implementation, the integral cumulative value is updated based on the concentration of toxic and harmful gases in the target area at the current moment, including: Update the gas concentration sequence based on the current concentration of toxic and harmful gases in the target area; The updated gas concentration sequence is convolved with a preset monotonically decreasing time decay weight function to obtain the updated cumulative integral value.

[0032] In this embodiment, the system maintains a fixed-length historical gas concentration sequence in memory, which stores a preset time window from the past. Continuously collected concentration data of toxic and harmful gases. When new concentration data for the current moment... Upon arrival, the system appends the data to the end of the sequence and removes the earliest data point from the beginning of the sequence, keeping the total sequence length constant, thus completing the rolling update of the concentration sequence. Let the historical time be... The current time is The continuous form of the monotonically decreasing time decay weight function is defined as follows: The exponentially decaying function is the natural form of the impulse response of a linear system with constant coefficients, and its decay constant is... It has a clear physical meaning—characterizing the rate at which the human body metabolizes and eliminates a specific toxic or harmful gas, expressed as the reciprocal of time. When When a larger value is taken, the weight of earlier exposures decays rapidly, indicating that the gas is metabolized quickly in the body, and the current toxic load mainly depends on recent exposure; when Smaller values ​​indicate that the effects of historical exposure last longer, suggesting that the gas tends to accumulate in the body. This modeling approach mathematically characterizes the natural laws of human physiological metabolism, allowing the integral accumulation value to reflect the actual cumulative biological effects. The updated gas concentration sequence... Compared with the aforementioned time decay weighting function in historical periods The formula for calculating the continuous form of the convolution integral is as follows: In the actual discrete sampling conditions of a digital monitoring system, the sensor operates at fixed time intervals. The output data, specifically the continuous convolution integral mentioned above, needs to be converted into a finite summation form to facilitate real-time execution by the embedded processor. Let the total number of sampling points within the preset historical time period be... The formula for calculating the cumulative value of discrete integrals is: The physical meaning of each term in this discrete expression is: the past... Concentration values ​​at each sampling time Multiply by the corresponding exponential decay weight Multiply by the sampling interval Cumulative results The estimated value of the toxic load in the body of workers at the current moment reflects both the level of exposure concentration and the duration of exposure. At the same time, it reflects the dynamic process of human metabolic clearance through exponential decay, thus providing a mathematical tool that conforms to physiological laws for the quantitative assessment of long-term low-concentration exposure risk.

[0033] In one possible implementation, the comprehensive risk index at the current moment is calculated based on the comprehensive risk index of the previous moment, the updated cumulative integral value, and the joint anomaly index at the current moment, including: The instantaneous evolution rate of the environmental state is obtained by performing first-order difference calculation on the joint anomaly index of the current moment and multiple historical moments. The instantaneous evolution rate is input into a preset mapping function to obtain dynamically updated weight coefficients; Using weighting coefficients, the comprehensive risk warning index of the previous time step, the updated cumulative integral value, and the joint anomaly index of the current time step are weighted and summed to obtain the comprehensive risk index of the current time step.

[0034] In this embodiment, the instantaneous evolution rate of the environmental state is obtained through first-order difference calculation. Let the joint anomaly index at the current moment be... The joint anomaly index at the previous sampling time was The sampling time interval is Then the formula for calculating the first difference is: This formula is essentially a numerical derivative of the time series. This represents the rate of change of the joint anomaly index over time. When... A positive value indicates that the environment is evolving in a more dangerous direction; the larger the value, the faster the deterioration. A negative value indicates that the environment is recovering or improving. Considering that accidents in confined spaces often deteriorate rapidly within minutes or even tens of seconds, this evolution rate can capture abrupt changes that conventional steady-state monitoring cannot identify. The acquired instantaneous evolution rate is input into a preset nonlinear mapping function, which can take the form of a logistic curve with saturation characteristics. in The preset basic evolution rate threshold, whose physical meaning is the critical rate that triggers the weight to increase rapidly, is usually set based on the evolution rate distribution of historical accident precursor data in a limited space. This is a sensitivity adjustment parameter that controls the steepness of the mapping curve near the threshold. The logistic function has the following characteristics: when... much smaller hour, Approaching 0; when Much larger hour, Approaching 1; while nearby, A rapid nonlinear transition occurs. This S-shaped saturation characteristic is widely used in control theory and signal processing to implement soft threshold switching, preventing system jitter near the threshold due to noise interference, while ensuring that the weighting coefficients can quickly transition to high values ​​when the environment truly deteriorates. Ultimately, the system utilizes these dynamically generated weighting coefficients. Combined with preset smoothing weight coefficients corresponding to the risk state at the previous moment And the toxicity weighting coefficient of the corresponding gas integral cumulative value Perform a weighted summation operation: in This is the comprehensive risk warning index from the previous moment. This is the current accumulated points value after the update. Let be the joint anomaly index at the current moment. The mathematical structure of this linearly weighted combination is essentially a fusion of a first-order recursive filter and a linear discriminant function: This introduces inertia into the system, allowing the comprehensive risk index to evolve smoothly over time; The item introduces long-term toxicity factors; The weighting coefficients of the items are dynamically adjusted according to the rate of environmental degradation. The sum of the three... It is a comprehensive risk warning index for the current moment, which can take into account both chronic cumulative damage and sudden change risks.

[0035] In one possible implementation, risk warnings are issued for the target area based on the comprehensive risk index at the current moment, including: The risk level of the target area at the current moment is obtained by comparing the comprehensive risk index at the current moment with the preset classification threshold range. Based on the risk level of the target area at the current moment, warnings are issued through corresponding warning methods; these methods include controlling ground displays and / or mobile terminals to output warning prompts, triggering on-site audible and visual alarm devices, and generating evacuation control signals.

[0036] In this embodiment, the system will calculate and output the comprehensive risk warning index for the current moment. The values ​​are compared with the graded risk thresholds pre-stored in non-volatile memory. Specifically, the system sets the first-level security warning threshold as follows: The second-level extreme danger threshold is The current comprehensive risk warning index is The system constructs a discrete system response state mapping function based on the above parameters, which is mathematically expressed as: in, The setting is based on the statistical fluctuation range of the comprehensive risk index time series under normal operating conditions. It is usually taken as the mean plus three times the standard deviation to ensure that 99.7% of normal fluctuations will not trigger the warning falsely. This requires setting environmental parameters based on the combined constraints of toxic gas concentration, water level, and flow rate in confined space operation safety standards, combined with the risk index distribution in historical accident cases, using statistical quantile methods or extreme value theory, to ensure that the combination of environmental parameters approaches the recognized danger boundary when the threshold is reached. The range of semi-closed and semi-open values ​​is defined as the prevention and attention interval, which is greater than or equal to The numerical range is defined as the extremely dangerous range.

[0037] When the system response state function outputs When the comprehensive risk warning index falls into the prevention and attention range, the system determines that the flow velocity and gas coupling environment in the confined space is deteriorating. The main control unit generates a warning data packet with a specific abnormality identifier through the network communication interface, and pushes it synchronously to the display screen of the ground monitoring center and the mobile terminal carried by the on-site monitoring personnel through the wireless link, triggering the client software pop-up window and interface color change to execute the preventive warning prompt.

[0038] When the system response state function outputs When the comprehensive risk warning index enters the extremely dangerous range, the system determines that there is a fatal hidden danger of a large release of toxic gas or a large-scale water surge in pipelines inside the confined space. The kernel suspends other normal processes and triggers the highest priority hardware interrupt control, outputting a high-level drive command to the underlying digital relay module to forcibly close the power supply circuit of the on-site audible and visual alarm equipment. At the same time, it sends an evacuation control signal to the on-site escape guidance equipment through the serial bus to activate the personnel evacuation indication path, ensuring that precious evacuation time is gained for on-site personnel when extreme danger occurs.

[0039] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0040] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0041] Figure 2 A schematic diagram of the structure of the confined space operation risk identification and early warning device provided in an embodiment of the present invention is shown. For ease of explanation, only the parts related to the embodiment of the present invention are shown, and are described in detail below: like Figure 2 As shown, the confined space operation risk identification and early warning device 2 includes: The difference calculation module 21 is used to calculate the difference vector between the real-time environmental state vector of the target area and the reference environmental state vector; wherein, the environmental state vector includes the concentration of toxic and harmful gases, water level and water flow velocity; Anomaly index calculation module 22 is used to calculate Mahalanobis distance based on the difference vector and the benchmark covariance matrix, as the joint anomaly index at the current moment; The risk index calculation module 23 is used to update the cumulative integral value based on the concentration of toxic and harmful gases in the target area at the current time, and to calculate the comprehensive risk index at the current time based on the comprehensive risk index at the previous time, the updated cumulative integral value and the joint anomaly index at the current time. The risk warning module 24 is used to provide risk warnings for the target area based on the comprehensive risk index at the current moment.

[0042] In one possible implementation, the difference calculation module 21 is also used for: Before calculating the difference between the real-time environmental state vector of the target area and the baseline environmental state vector, the mathematical expectations of the concentration of toxic and harmful gases, water level and water flow velocity of the target area in historical periods are calculated respectively to form the baseline mean vector. Based on the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, calculate the covariance values ​​between parameter sequences of each dimension. A positive definite symmetric square matrix representing the joint distribution characteristics is constructed based on the values ​​of each covariance, serving as the benchmark covariance matrix.

[0043] In one possible implementation, the anomaly index calculation module 22 is specifically used for: Perform matrix multiplication operations on the transpose of the difference vector, the reference covariance matrix, and the difference vector in sequence to obtain a quadratic scalar that characterizes the degree of multi-parameter coupling distortion. The joint anomaly index at the current moment is obtained by taking the square root of the quadratic scalar.

[0044] In one possible implementation, the risk index calculation module 23 is specifically used for: Update the gas concentration sequence based on the current concentration of toxic and harmful gases in the target area; The updated gas concentration sequence is convolved with a preset monotonically decreasing time decay weight function to obtain the updated cumulative integral value.

[0045] In one possible implementation, the risk index calculation module 23 is specifically used for: The instantaneous evolution rate of the environmental state is obtained by performing first-order difference calculation on the joint anomaly index of the current moment and multiple historical moments. The instantaneous evolution rate is input into a preset mapping function to obtain dynamically updated weight coefficients; Using weighting coefficients, the comprehensive risk warning index of the previous time step, the updated cumulative integral value, and the joint anomaly index of the current time step are weighted and summed to obtain the comprehensive risk index of the current time step.

[0046] In one possible implementation, the risk warning module 24 is specifically used for: The risk level of the target area at the current moment is obtained by comparing the comprehensive risk index at the current moment with the preset classification threshold range. Based on the risk level of the target area at the current moment, warnings are issued through corresponding warning methods; these methods include controlling ground displays and / or mobile terminals to output warning prompts, triggering on-site audible and visual alarm devices, and generating evacuation control signals.

[0047] This invention constructs a three-dimensional environmental state vector containing toxic and harmful gas concentrations, water levels, and water flow velocities. Based on the degree of vector deviation from the baseline state and combined with the covariance relationship between multiple parameters, Mahalanobis distance is calculated, enabling a comprehensive quantification of the joint distortion degree of multiple environmental parameters. Compared to methods that rely on independent thresholds for single parameters, this approach can identify potential hazards in advance when a single parameter has not yet exceeded its limit, but the coordinated changes of multiple parameters have already shown an abnormal trend. Furthermore, by introducing a gas concentration integral accumulation value based on time decay weights, it can reflect the hidden harm suffered by workers due to prolonged exposure to low-concentration harmful gas environments. Combined with the risk status of the previous moment for dynamic weighted fusion, the risk assessment considers both sudden changes and cumulative effects, effectively improving the accuracy and timeliness of risk identification and early warning in confined spaces. Finally, a tiered early warning strategy triggers corresponding response actions at different risk levels, helping to provide more adequate safety guarantees for confined space operations while ensuring normal construction order.

[0048] Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Figure 3 As shown, the electronic device 3 of this embodiment includes a processor 30 and a memory 31. The memory 31 stores a computer program 32. When the processor 30 executes the computer program 32, it implements the steps in the various method embodiments described above. Alternatively, when the processor 30 executes the computer program 32, it implements the functions of each module / unit in the various device embodiments described above.

[0049] For example, computer program 32 may be divided into one or more modules / units, which are stored in memory 31 and executed by processor 30 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 32 in electronic device 3.

[0050] Electronic device 3 may include, but is not limited to, processor 30 and memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 3 may also include input / output devices, network access devices, buses, etc.

[0051] The processor 30 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0052] The memory 31 can be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. The memory 31 can also be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. Furthermore, the memory 31 can include both internal and external storage units of the electronic device 3. The memory 31 is used to store the computer program 32 and other programs and data required by the electronic device 3. The memory 31 can also be used to temporarily store data that has been output or will be output.

[0053] For the sake of simplicity and clarity, only the above-described functional modules / units are used as examples. In practical applications, the functions described above can be assigned to different functional modules / units as needed. These modules / units can be implemented in hardware, software, or a combination of both.

[0054] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the methods described in the above-described method embodiments.

[0055] This invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the methods described in the above-described method embodiments.

[0056] Computer programs include computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. Computer-readable media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0057] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0058] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for identifying and warning of operational risks in a confined space, characterized in that, include: Calculate the difference vector between the real-time environmental state vector of the target area and the baseline environmental state vector; whereby the environmental state vector includes the concentration of toxic and harmful gases, water level, and water flow velocity; The Mahalanobis distance is calculated based on the difference vector and the benchmark covariance matrix, and used as the joint anomaly index at the current moment. The integral cumulative value is updated based on the concentration of toxic and harmful gases in the target area at the current time, and the comprehensive risk index at the current time is calculated based on the comprehensive risk index at the previous time, the updated integral cumulative value, and the joint anomaly index at the current time. Risk warnings are issued for the target area based on the comprehensive risk index at the current moment.

2. The method for identifying and warning of operational risks in confined spaces according to claim 1, characterized in that, Before calculating the difference between the real-time environmental state vector of the target area and the reference environmental state vector, the method further includes: Calculate the mathematical expectations of the concentration of toxic and harmful gases, water level and water flow velocity in the target area during historical periods, and form a benchmark mean vector; Based on the concentration of toxic and harmful gases, water level, and water flow velocity in the target area during historical periods, calculate the covariance values ​​between the parameter sequences of each dimension. A positive definite symmetric square matrix representing the joint distribution characteristics is constructed based on the values ​​of each covariance, serving as the benchmark covariance matrix.

3. The method for identifying and warning of operational risks in confined spaces according to claim 1, characterized in that, The calculation of the Mahalanobis distance based on the difference vector and the benchmark covariance matrix, as the joint anomaly index at the current moment, includes: The transpose of the difference vector, the reference covariance matrix, and the difference vector are sequentially multiplied to obtain a quadratic scalar that characterizes the degree of multi-parameter coupling distortion. The square root operation is performed on the quadratic scalar to obtain the joint anomaly index at the current moment.

4. The method for identifying and warning of operational risks in confined spaces according to claim 1, characterized in that, The step of updating the integral cumulative value based on the concentration of toxic and harmful gases in the target area at the current moment includes: Update the gas concentration sequence based on the current concentration of toxic and harmful gases in the target area; The updated gas concentration sequence is convolved with a preset monotonically decreasing time decay weight function to obtain the updated cumulative integral value.

5. The method for identifying and warning of operational risks in confined spaces according to claim 1, characterized in that, The calculation of the comprehensive risk index at the current moment, based on the comprehensive risk index of the previous moment, the updated cumulative integral value, and the joint anomaly index at the current moment, includes: The instantaneous evolution rate of the environmental state is obtained by performing first-order difference calculation on the joint anomaly index of the current moment and multiple historical moments. The instantaneous evolution rate is input into a preset mapping function to obtain dynamically updated weight coefficients; Using the aforementioned weighting coefficients, the comprehensive risk warning index of the previous moment, the updated cumulative integral value, and the joint anomaly index of the current moment are weighted and summed to obtain the comprehensive risk index of the current moment.

6. The method for identifying and warning of operational risks in confined spaces according to claim 1, characterized in that, The risk warning for the target area based on the comprehensive risk index at the current moment includes: The risk level of the target area at the current moment is obtained by comparing the comprehensive risk index at the current moment with the preset classification threshold range. Based on the risk level of the target area at the current moment, an early warning is issued through a corresponding warning method; wherein, the early warning method includes controlling the ground display screen and / or mobile terminal to output early warning prompts, triggering on-site audible and visual alarm devices and generating evacuation control signals.

7. A confined space operation risk identification and early warning device, characterized in that, include: The difference calculation module is used to calculate the difference vector between the real-time environmental state vector of the target area and the baseline environmental state vector; wherein, the environmental state vector includes the concentration of toxic and harmful gases, water level and water flow velocity; Anomaly index calculation module is used to calculate Mahalanobis distance based on the difference vector and the benchmark covariance matrix, as the joint anomaly index at the current moment; The risk index calculation module is used to update the cumulative integral value based on the concentration of toxic and harmful gases in the target area at the current time, and to calculate the comprehensive risk index at the current time based on the comprehensive risk index at the previous time, the updated cumulative integral value, and the joint anomaly index at the current time. The risk warning module is used to issue risk warnings to the target area based on the comprehensive risk index at the current moment.

8. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images