Consistency regulation and control method for combustible gas alarm device
By injecting calibration gas into combustible gas alarm devices and performing environmental normalization, calculating the group dispersion index, and dynamically adjusting the weights for compensation, the problems of device response characteristic divergence and high operation and maintenance costs are solved, and consistency evaluation and self-healing closed loop among devices are realized.
Patent Information
- Application Number
- CN202511149338.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-18
- Publication Date
- 2025-11-21
AI Technical Summary
Existing combustible gas alarm devices suffer from high false alarm rates due to the divergence in response characteristics among devices caused by environmental fluctuations in high-density deployment scenarios. They also lack quantitative monitoring of the dispersion of group performance. Traditional compensation techniques cannot adapt to nonlinear aging, resulting in high operation and maintenance costs and low system availability.
By periodically injecting calibration gas into the acquisition device to collect signals and environmental parameters, and using the normalized response speed of the environmental parameters, the population dispersion index is calculated, and weights are dynamically allocated for alarm compensation, forming a performance self-healing closed loop.
It enables consistency assessment among devices under complex operating conditions, actively captures drift of individual devices, reduces false alarm rate and maintenance costs, and improves the real-time performance and availability of the system.
Smart Images
Figure CN120997993A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of safety monitoring technology, specifically to a method for consistent control of combustible gas alarm devices. Background Technology
[0002] In the field of industrial combustible gas safety monitoring, the reliability of multi-detector collaborative alarm systems highly depends on the consistency of output signals among devices. However, existing technologies suffer from three significant drawbacks: First, traditional calibration methods rely on single-point static calibration, which cannot eliminate the dynamic interference of actual complex operating conditions (such as temperature gradients in oil and gas pipelines or high humidity environments in chemical plants) on the sensor baseline. This leads to continuous divergence in the response characteristics of devices of the same model operating at different locations, and the false alarm rate increases with environmental fluctuations. Second, there is a lack of a quantitative monitoring mechanism for the dispersion of group performance. Most systems only set fixed alarm thresholds, which cannot capture early performance drift of individual devices, nor can they identify systemic risks such as the aging of a batch of devices. Intervention is often reactive after a failure occurs, resulting in delayed handling of safety hazards. In addition, existing compensation techniques mostly use linear offset correction or manual reset calibration. The compensation process requires interruption of monitoring and cannot adapt to the nonlinear aging characteristics of devices. Especially in accelerated aging environments such as high temperature and high humidity, frequent offline calibration significantly reduces system availability. These drawbacks collectively lead to a high false alarm rate in high-density deployment scenarios, and the maintenance costs are mostly spent on manual calibration, which seriously restricts the promotion of large-scale intelligent alarm systems. Summary of the Invention
[0003] To maintain consistency among combustible gas alarm devices in high-density deployment scenarios and reduce manual maintenance costs, this invention proposes a consistency control method for combustible gas alarm devices, comprising the following steps:
[0004] S1: In the actual operating environment, the concentration output signals and environmental parameters of multiple combustible gas alarm devices of the same model are periodically collected by injecting calibration gas;
[0005] S2: Calculate the response speed parameter of each device based on the concentration output signal. This response speed parameter is the time required for the concentration output signal to rise from the baseline value to the target value when the gas concentration reaches the preset threshold.
[0006] S3: Normalize the response speed parameters of the corresponding devices using environmental parameters;
[0007] S4: Calculate the population dispersion index characterizing the differences between devices based on the normalized response speed parameter set;
[0008] S5: When the dispersion index of the group exceeds the preset threshold for a period of time that exceeds the preset warning time, a consistency control instruction is generated;
[0009] S6: Execute the consistency control command to perform alarm compensation under dynamic weight adjustment for individual devices identified as having exceeded the dispersion index limit.
[0010] This invention utilizes environmental parameters to dynamically normalize response speed parameters, effectively isolating the interference of complex operating conditions such as temperature, humidity, and air pressure on the baseline performance of devices. It overcomes the shortcomings of traditional single-point calibration which is affected by environmental fluctuations, enabling devices at different locations to evaluate performance under a unified benchmark. At the same time, by using a group discrete quantification index, it continuously monitors the dispersion of response speeds among a group of devices of the same model, thereby achieving proactive capture of early performance drift of individual devices and solving the problem of lag in existing technologies that rely on post-event manual inspection.
[0011] Furthermore, in step S3, the normalization process is implemented through the following steps:
[0012] Obtain the difference between each environmental parameter and the corresponding preset benchmark environmental parameter, and multiply the difference by the corresponding environmental compensation coefficient to obtain the compensation item for each environmental parameter;
[0013] The comprehensive compensation factor is obtained by multiplying the compensation terms of each environmental parameter, and the normalized response rate parameter is obtained by dividing the original response rate parameter by the comprehensive compensation factor.
[0014] Furthermore, in step S4, the group dispersion index is the coefficient of variation of the response speed parameter. The coefficient of variation characterizes the degree of dispersion of the response speed of a group of devices of the same model. It is calculated as the ratio of the standard deviation to the average value of the response speed parameters of all combustible gas alarm devices.
[0015] Furthermore, in step S6, the alarm compensation under the dynamic weight allocation adjustment specifically includes the following operations:
[0016] Extract the historical dataset {t} of individual devices identified as exceeding the dispersion index within a time window T. i ,v i}, where t i Let v be the timestamp of the i-th data collection. i This represents the original output signal of the device during the i-th data acquisition.
[0017] The original output signal v is calibrated using the device's calibration function. i Concentration estimate converted to the i-th data acquisition
[0018] Based on the reference concentration value c of the i-th data point i Compared with concentration estimates The deviation and data timeliness are calculated to dynamically allocate the weight w for the i-th data point. i ;
[0019] Based on dynamically assigned weights w i For reference concentration c i The weighted output value is used to generate the compensation value.
[0020] Furthermore, the dynamically allocated weight w i Calculated using the following formula:
[0021]
[0022] In the formula, α and β are weighting coefficients, γ is the time decay coefficient, and Δt i For the current time and t i The time difference.
[0023] Furthermore, the concentration output signal is compensated using the following formula:
[0024]
[0025] In the formula, c out The output signal is the compensated concentration signal, where i is the data point index, n is the total number of data points within the time window T, k is the discrete adaptive compensation gain coefficient, and c... now This is the reference concentration value at the current moment. The reference concentration value c within the time window T i The average value.
[0026] Furthermore, the reference concentration value c i The known concentration value of the injected calibration gas.
[0027] Furthermore, following step S6, the following step is also included:
[0028] S61: Reacquire the response speed parameters of the compensation device, normalize them, and calculate the deviation of the device from the population median.
[0029] S62: If the deviation is greater than the preset deviation, inject calibration gas of stepped concentration into the device; otherwise, update the weighting coefficient of the device.
[0030] S63: Reconstruct the device calibration function based on the concentration output signals of each step concentration and write it into the device memory.
[0031] Compared with the prior art, the present invention has at least the following beneficial effects:
[0032] (1) The present invention proposes a consistency control method for combustible gas alarm devices, which uses environmental parameters to dynamically normalize the response speed parameters, effectively isolates the interference of complex working conditions such as temperature, humidity and air pressure on the baseline performance of the device, overcomes the defect of traditional single-point calibration being affected by environmental fluctuations, and enables devices at different locations to evaluate their performance under a unified benchmark.
[0033] (2) By introducing the coefficient of variation as a quantitative indicator of the discreteness of the group, the early performance drift of individual devices is actively captured by continuously monitoring the dispersion of the response speed of the same type of device group, thus solving the problem of the lag of the existing technology that relies on manual investigation after the fact.
[0034] (3) When the dispersion exceeds the limit, the compensation mechanism is triggered by multi-dimensional optimization: based on the dynamic weight allocation of time decay coefficient and concentration deviation, historical effective data is accurately extracted, and the nonlinear drift is corrected in real time by the discrete adaptive compensation gain coefficient, and the device output calibration is automatically completed without interrupting the monitoring.
[0035] (4) By reconstructing the step concentration calibration function and iteratively updating the weight coefficient, a self-healing closed loop for device performance degradation is formed, reducing the cost of manual maintenance. Attached Figure Description
[0036] Figure 1 This is a flowchart illustrating the steps of a consistency control method for a combustible gas alarm device. Detailed Implementation
[0037] The following are specific embodiments of the present invention, which are described in conjunction with the accompanying drawings. However, the present invention is not limited to these embodiments.
[0038] Although the factory calibration technology for combustible gas alarm devices is relatively mature, it still faces severe challenges in long-term operation. Existing solutions struggle to effectively overcome the differential impact of dynamic fluctuations in environmental parameters (such as changes in temperature, humidity, and air pressure) on the response characteristics of a group of devices, leading to systematic deviations in the output signals of devices of the same model at the same gas concentration over time. Furthermore, the lack of a continuous quantitative evaluation mechanism for the performance dispersion of the device group makes it impossible to promptly identify abnormal drift trends in individual devices, often requiring passive waiting for a fault to occur before manual intervention. More importantly, traditional compensation methods rely on fixed thresholds or linear correction models, which not only fail to adapt to the nonlinear aging characteristics of devices but also require interruption of monitoring and manual injection of calibration gas during the compensation process, severely restricting system real-time performance. These shortcomings cause high-density deployed alarm systems to gradually lose group consistency under complex operating conditions. Therefore, such as... Figure 1 As shown, this invention proposes a consistency control method for combustible gas alarm devices, comprising the following steps:
[0039] S1: In the actual operating environment, the concentration output signals and environmental parameters of multiple combustible gas alarm devices of the same model are periodically collected by injecting calibration gas;
[0040] S2: Calculate the response speed parameter of each device based on the concentration output signal. This response speed parameter is the time required for the concentration output signal to rise from the baseline value to the target value when the gas concentration reaches the preset threshold.
[0041] S3: Normalize the response speed parameters of the corresponding devices using environmental parameters;
[0042] S4: Calculate the population dispersion index characterizing the differences between devices based on the normalized response speed parameter set;
[0043] S5: When the dispersion index of the group exceeds the preset threshold for a period of time that exceeds the preset warning time, a consistency control instruction is generated;
[0044] S6: Execute the consistency control command to perform alarm compensation under dynamic weight adjustment for individual devices identified as having exceeded the dispersion index limit.
[0045] Specifically, in actual operating environments, this method involves periodically injecting a specific concentration of calibration gas (such as methane or propane) into the monitoring area while simultaneously collecting key data from combustible gas alarm devices of the same batch. In practice, the calibration gas is injected according to a preset concentration gradient (e.g., low, medium, and high concentrations in stages), with each injection cycle continuing until the device's output signal completes the entire process from baseline to a steady state, thus fully recording the dynamic response curve of the concentration output signal. Data acquisition covers two core parameters: first, the device's raw electrical signal output (usually voltage or current values); and second, real-time environmental parameters (including temperature, humidity, air pressure, and optional wind speed data). The acquisition of environmental parameters is strictly synchronized with the gas signal acquisition to ensure the accuracy of subsequent environmental compensation calculations. The calibration gas triggering cycle employs a dual-mode design: under normal monitoring, a single concentration of gas is automatically injected at fixed intervals (e.g., 24 hours); when the system detects a performance anomaly, it switches to a stepped concentration injection mode (e.g., 10%, 30%, and 50% lower explosive limit concentration) to obtain the device's full-range response characteristics. All collected data is stored with timestamps, providing a basic dataset for subsequent response speed analysis, environmental normalization processing, and calibration function reconstruction.
[0046] After data acquisition, the core task of step S2 is to quantify and extract the response speed parameters of each device from the dynamic response curve. First, the raw output signal is preprocessed: high-frequency noise interference is eliminated through low-pass filtering, and signal fluctuations are smoothed using a sliding window algorithm. Then, the baseline value is determined as the average signal value under stable conditions before calibration gas injection (usually the arithmetic mean of the data from the first 10 seconds before injection), while the target value is set as the theoretical output signal value corresponding to when the calibration gas concentration reaches a preset threshold. The precise calculation of the response speed parameters is achieved through a time-domain feature extraction algorithm, starting timing from the point when the signal first exceeds the baseline value (response start point) and ending timing when the signal remains stable within a certain range of the target value for 0.5 seconds (response end point). This time interval is the response speed parameter. For multi-device parallel processing scenarios, the control system employs distributed timestamp alignment technology to ensure a unified start and end time determination benchmark for all devices. Specifically, when there is stepped concentration injection data, the response speed parameters for each concentration step are calculated separately, forming a device response speed-concentration characteristic matrix, providing multi-dimensional features for subsequent performance degradation analysis.
[0047] After obtaining the original response speed parameters of each device, it is necessary to eliminate the interference of environmental fluctuations on performance evaluation and achieve device comparability across time and space conditions. This invention uses a pre-stored set of benchmark environmental parameters (including standard temperature, relative humidity, atmospheric pressure, and other laboratory calibration benchmark values) in a database to calculate the deviation between the currently collected environmental parameters and the corresponding benchmark values in real time. Subsequently, each environmental parameter is multiplied by a device-specific compensation coefficient (determined through factory calibration tests: temperature, humidity, and pressure parameters are independently adjusted in a controlled environment chamber, quantifying the impact of unit changes in each parameter on the response speed and fitting it to the slope of a linear equation), generating temperature compensation, humidity compensation, and atmospheric pressure compensation terms. The key operation lies in the dynamic synthesis of a comprehensive compensation factor: by adding 1 to each environmental compensation term and then multiplying them together, i.e.:
[0048] Comprehensive factor = (1+λ) t ·ΔT)×(1+λ h ·ΔH)×(1+λ p ·ΔP),
[0049] Where λ t , λ h , λ p These are the temperature / humidity / pressure compensation coefficients, respectively. Finally, the original response rate parameters are divided by this comprehensive compensation factor to output the normalized response rate parameters.
[0050] Based on the normalized response speed parameter set, the performance consistency level of a group of devices of the same model can be quantitatively evaluated. Specifically, the dataset is first screened for validity: outlier data points during periods when environmental parameters exceed limits are removed. Then, the group dispersion index is calculated—defined as the coefficient of variation of the normalized response speed parameters. The calculation process strictly follows statistical norms: first, the arithmetic mean of all valid device response speed parameters is calculated; then, the standard deviation of each parameter from the mean is calculated; finally, the standard deviation is divided by the mean and converted to a percentage for output. To improve the robustness of the evaluation, the control system also introduces a triple verification mechanism: when a single device parameter deviates from the group median by more than three times the standard deviation, the response curve of that device is automatically checked (comparing the similarity between the original signal waveform and the calibration template). Only after confirmation that it is not a misjudgment can it be included in the dispersion calculation. For multi-concentration level scenarios, sub-coefficients of variation are calculated for low / medium / high concentrations, and the largest sub-coefficient is used as the final group dispersion index. The dispersion calculation results are mapped to the dynamic baseline model in real time: the system records the coefficient of variation sequence over 30 consecutive periods and updates the dispersion threshold benchmark through an exponential smoothing algorithm to avoid false alarms caused by fixed thresholds. The final generated population dispersion index is simultaneously labeled with a confidence level.
[0051] Subsequently, based on the obtained population dispersion index, the control system first compares the real-time calculated coefficient of variation with a dynamic threshold: this threshold is composed of a baseline value plus an environmental adaptive fluctuation. The judgment logic adopts a dual time constraint mechanism—population consistency disorder is judged only when the dispersion index of three consecutive collection cycles exceeds the threshold and the cumulative over-limit duration is greater than the preset warning duration. To prevent false triggering caused by instantaneous interference, the system simultaneously initiates multi-dimensional verification: checking whether the fluctuation of environmental parameters during the same period is within a stable range, and verifying whether the geographical distribution of over-limit devices shows clustering (such as activating regional mode alarms when the proportion of over-limit devices in the same partition is >40%). If the verification passes, a hierarchical control instruction is generated: Level 1 instruction (dispersion over-limit <10%) only marks the devices to be observed and reduces their data weight; Level 2 instruction (over-limit 10-20%) triggers background compensation pre-calculation; Level 3 instruction (over-limit >20%) immediately pushes an alarm to the operation and maintenance terminal and initiates the real-time compensation process. All instructions are automatically associated with environmental parameter tags when generated, providing operating condition context for subsequent compensation. Finally, the system records the characteristic fingerprint of this event (including dispersion curve, environmental parameter spectrum, and device distribution heat map) and writes it into the historical case library for the self-learning model to optimize the threshold strategy.
[0052] After generating a consistent control command, in order to perform precise adaptive alarm compensation for abnormal devices, this invention first locates the individual device with excessive dispersion and extracts the historical dataset within the most recent time window T from its local memory, including the timestamp t. i With the original output signal vi The timing pair. The original signal v is converted by calling the device's calibration function (usually a factory-stored second-order polynomial used to convert the electrical signal into a corresponding gas concentration signal). i Convert to concentration estimate
[0053] The core of the compensation algorithm lies in the dynamic allocation of weights: for each data point, weights are allocated based on its concentration deviation. (where c) i (Known concentration of the injected calibration gas) and data timeliness (current time versus t) i The difference Δt i The dynamic weight allocation is calculated using the following formula:
[0054]
[0055] Where α and β are weighting coefficients. The time decay coefficient γ in the formula causes the weight of recent data to increase exponentially, while the concentration bias term, with its inverse weighting, suppresses interference from outlier data. The final compensated output value is generated by a dual-channel fusion algorithm: the main channel is based on the weighted concentration values. Generate a baseline compensation amount; the auxiliary channel introduces a discrete adaptive gain coefficient k (k = 0.5 × current dispersion index) to adjust the real-time deviation. To perform magnification correction, that is:
[0056]
[0057] Among them, c out The output signal is the compensated concentration signal, where i is the data point index, n is the total number of data points within the time window T, k is the discrete adaptive compensation gain coefficient, and c... now This is the reference concentration value at the current moment. The reference concentration value c within the time window T i The average value.
[0058] After compensation is completed, closed-loop verification is automatically triggered: the response speed parameters of the device are reacquired. If the deviation from the population median is still >5%, the deep maintenance mode is started - a step concentration gas of 10% / 30% / 50% LEL is injected into the device, the calibration function is reconstructed based on the full range response data and burned into the device memory to complete the characteristic self-calibration.
[0059] In summary, the present invention proposes a consistency control method for combustible gas alarm devices. This method utilizes environmental parameters to dynamically normalize the response speed parameters, effectively isolating the interference of complex operating conditions such as temperature, humidity, and air pressure on the baseline performance of the device. It overcomes the shortcomings of traditional single-point calibration which is affected by environmental fluctuations, enabling devices at different locations to be evaluated under a unified benchmark.
[0060] By introducing the coefficient of variation as a quantitative indicator of population discreteness, and by continuously monitoring the dispersion of response speed among a group of devices of the same model, the early performance drift of individual devices is actively captured, solving the problem of lag in existing technologies that rely on manual investigation after the fact.
[0061] A compensation mechanism triggered by exceeding dispersion limits integrates multi-dimensional optimization: based on dynamic weight allocation of time decay coefficient and concentration deviation, historical effective data is accurately extracted, and combined with discrete adaptive compensation gain coefficient to correct nonlinear drift in real time, automatically completing device output calibration without interrupting monitoring. Furthermore, through the reconstruction of a stepped concentration calibration function and iterative updating of weight coefficients, a self-healing closed loop for device performance degradation is formed, reducing manual maintenance costs.
[0062] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.
[0063] Furthermore, in this invention, descriptions involving terms such as "first," "second," and "a" are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0064] In this invention, unless otherwise explicitly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.
[0065] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.
Claims
1. A method for consistent control of a combustible gas alarm device, characterized in that, Including the following steps: S1: In the actual operating environment, the concentration output signals and environmental parameters of multiple combustible gas alarm devices of the same model are periodically collected by injecting calibration gas; S2: Calculate the response speed parameter of each device based on the concentration output signal. This response speed parameter is the time required for the concentration output signal to rise from the baseline value to the target value when the gas concentration reaches the preset threshold. S3: Normalize the response speed parameters of the corresponding devices using environmental parameters; S4: Calculate the population dispersion index characterizing the differences between devices based on the normalized response speed parameter set; S5: When the dispersion index of the group exceeds the preset threshold for a period of time that exceeds the preset warning time, a consistency control instruction is generated; S6: Execute the consistency control command to perform alarm compensation under dynamic weight adjustment for individual devices identified as having exceeded the dispersion index limit.
2. The consistency control method for a combustible gas alarm device as described in claim 1, characterized in that, In step S3, the normalization process is implemented through the following steps: Obtain the difference between each environmental parameter and the corresponding preset benchmark environmental parameter, and multiply the difference by the corresponding environmental compensation coefficient to obtain the compensation item for each environmental parameter; The comprehensive compensation factor is obtained by multiplying the compensation terms of each environmental parameter, and the normalized response rate parameter is obtained by dividing the original response rate parameter by the comprehensive compensation factor.
3. The consistency control method for a combustible gas alarm device as described in claim 1, characterized in that, In step S4, the group dispersion index is the coefficient of variation of the response speed parameter. The coefficient of variation characterizes the degree of dispersion of the response speed of a group of devices of the same model. It is calculated as the ratio of the standard deviation to the average value of the response speed parameters of all combustible gas alarm devices.
4. The consistency control method for a combustible gas alarm device as described in claim 1, characterized in that, In step S6, the alarm compensation under dynamic weight adjustment specifically includes the following operations: Extract the historical dataset {t} of individual devices identified as exceeding the dispersion index within a time window T. i ,v i }, where t i Let v be the timestamp of the i-th data collection. i This represents the original output signal of the device during the i-th data acquisition. The original output signal v is calibrated using the device's calibration function. i Concentration estimate converted to the i-th data acquisition Based on the reference concentration value c of the i-th data point i Compared with concentration estimates The deviation and data timeliness are calculated to dynamically allocate the weight w for the i-th data point. i ; Based on dynamically assigned weights w i For reference concentration c i The weighted output value is used to generate the compensation value.
5. The consistency control method for a combustible gas alarm device as described in claim 4, characterized in that, The dynamically allocated weight w i Calculated using the following formula: In the formula, α and β are weighting coefficients, γ is the time decay coefficient, and Δt i For the current time and t i The time difference.
6. The consistency control method for a combustible gas alarm device as described in claim 4, characterized in that, The concentration output signal is compensated using the following formula: In the formula, c out The output signal is the compensated concentration signal, where i is the data point index, n is the total number of data points within the time window T, k is the discrete adaptive compensation gain coefficient, and c... now This is the reference concentration value at the current moment. The reference concentration value c within the time window T i The average value.
7. A consistency control method for a combustible gas alarm device as described in any one of claims 4 to 6, characterized in that, The reference concentration value c i The known concentration value of the injected calibration gas.
8. The consistency control method for a combustible gas alarm device as described in claim 1, characterized in that, Following step S6, the following step is also included: S61: Reacquire the response speed parameters of the compensation device, normalize them, and calculate the deviation of the device from the population median. S62: If the deviation is greater than the preset deviation, inject calibration gas of stepped concentration into the device; otherwise, update the weighting coefficient of the device. S63: Reconstruct the device calibration function based on the concentration output signals of each step concentration and write it into the device memory.